Enzymatically crosslinked silk fibroin-based hierarchical scaffolds for

Subscriber access provided by United Arab Emirates University | Libraries Deanship

Biological and Medical Applications of Materials and Interfaces

Enzymatically crosslinked silk fibroin-based hierarchical scaffolds for osteochondral regeneration Viviana Ribeiro, Sandra Pina, João Bebiano Costa, Ibrahim Fatih Cengiz, Luis García-Fernández, María del Mar Fernández-Gutiérrez, Olga C. Paiva, Ana Leite Oliveira, Julio San Roman, Joaquim M. Oliveira, and Rui L. Reis ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b21259 • Publication Date (Web): 04 Jan 2019 Downloaded from http://pubs.acs.org on January 8, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Enzymatically crosslinked silk fibroin-based hierarchical scaffolds for osteochondral regeneration Viviana P. Ribeiro1,2*, Sandra Pina1,2, João B. Costa1,2, Ibrahim Fatih Cengiz1,2, Luis García-Fernández3,4, Maria del Mar Fernández-Gutiérrez3,4, Olga C. Paiva5, Ana L. Oliveira6, Julio San-Román3,4, Joaquim M. Oliveira1,2,7 and Rui L. Reis1,2,7 13B’s

Research Group, I3Bs - Research Institute on Biomaterials, Biodegradables and Biomimetics of University of Minho, Headquarters of the European Institute of Excellence on Tissue Engineering and Regenerative Medicine, Avepark, Parque de Ciência e Tecnologia, Zona Industrial da Gandra, 4805-017 Barco, Guimarães, Portugal; 2ICVS/3B’s - PT Government Associate Laboratory, Braga/Guimarães, 4805-017, Portugal; 3Institute of Polymer Science and Technology, Polymeric Nanomaterials and Biomaterials Department, Spanish Council for Scientific Research (ICTP-CSIC), 28006 Madrid, Spain; 4Centro de Investigación Biomédica en Red. Bioingeniería, Biomateriales y Nanomedicina (CIBER-BBN), 28029 Madrid, Spain; 5ISEP - School of Engineering, Polytechnic Institute of Porto, 4200-072 Porto, Portugal; 6CBQF – Centro de Biotecnologia e Química Fina, Laboratório Associado, Escola Superior de Biotecnologia, Universidade Católica Portuguesa, 4200-072 Porto, Portugal; 7The Discoveries Centre for Regenerative and Precision Medicine, Headquarters at University of Minho, Avepark, 4805-017 Barco, Guimarães, Portugal.

*Corresponding

author:

Viviana P. Ribeiro 3B’s Research Group I3Bs - Research Institute on Biomaterials, Biodegradables and Biomimetics of University of Minho Headquarters of the European Institute of Excellence on Tissue Engineering and Regenerative Medicine Tel: +351-253-510900 Fax: +351-253-510909 ORCID: 0000-0002-3679-0759

Keywords Horseradish peroxidase-mediated crosslinking; Silk fibroin; Ion-doped β-Tricalcium phosphate; Bilayered scaffold; Hierarchical structure; Osteochondral tissue engineering

ACS Paragon Plus Environment

1

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 51

Abstract Osteochondral (OC) regeneration has been facing several limitations in orthopedic surgery, owing the complexity of the OC tissue that simultaneously entails the restoration of articular cartilage and subchondral bone diseases. In this study, novel biofunctional hierarchical scaffolds composed of a horseradish peroxidase (HRP)-crosslinked silk fibroin (SF) cartilage-like layer (HRP-SF layer) fully integrated into a HRP-SF/ZnSr-doped β-tricalcium phosphate (β-TCP) subchondral bone-like layer (HRP-SF/dTCP layer), were proposed as a promising strategy for OC tissue regeneration. For comparative purposes, a similar bilayered structure produced with no ion incorporation (HRP-SF/TCP layer) was used. A homogeneous porosity distribution was achieved throughout the scaffolds as shown by micro-computed tomography analysis. The ion-doped bilayered scaffolds presented a wet compressive modulus (226.56 ± 60.34 kPa) and dynamic mechanical properties (ranging from 403.56 ± 111.62 to 593.56 ± 206.90 kPa) superior to that of the control bilayered scaffolds (189.18 ± 90.80 kPa and ranging from 262.72 ± 59.92 to 347.68 ± 93.37 kPa, respectively). The apatite-crystals formation, after immersion in simulated body fluid (SBF) was observed in the subchondral bone-like layers for the scaffolds incorporating TCP powders. Human osteoblasts (hOBs) and human articular chondrocytes (hACs) were co-cultured onto the bilayered structures, and monocultured in the respective cartilage and subchondral bone-half of the partitioned scaffolds. Both cell types showed good adhesion and proliferation in the scaffold compartments, as well as, adequate integration of the interface regions. Osteoblasts produced a mineralized extracellular matrix (ECM) in the subchondral bone-like layers, and chondrocytes showed GAGs deposition. Gene expression profile was different in the distinct zones of the bilayered constructs and the intermediate regions showed pre-hypertrophic chondrocytes gene expression, especially on the BdTCP constructs. Immunofluorecence analysis have supported these observations. This study, showed that the proposed bilayered scaffolds allowed a specific stimulation of the chondrogenic and osteogenic cells in the co-culture system together with the formation of an osteochondral-like tissue interface. Hence, the structural adaptability, suitable mechanical properties, and biological performance of the hierarchical scaffolds make these constructs a desired strategy for OC defects regeneration.


2

Page 3 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. Introduction Osteochondral (OC) defects are joint injuries that affect simultaneously the articular cartilage and the underlying subchondral bone 1. The causes associated with these deformities can be of natural degradation or trauma related injuries. The natural wear of articular cartilage tissue is commonly observed in an aging population and usually leads to degenerative osteoarthritis (OA), the biggest cause of natural OC defects 2. The prevalence of symptomatic osteoarthritis in adults ranges from about 7 to 17%, increasing at ages superior to 65 years old 3, being higher among women than men. Knee (femoral-tibial joint) OC defects are very common in athletes of high competition sports involving high compressive load-bearing of the articular region 4. The treatments of OC defects include arthroscopic debridement, the use of autografts or allografts, and bone marrow stimulation technique 5. Nevertheless, these are considered palliative temporary solutions that can present limitations of immune rejection, disease transmission, and donor site morbidity 6. Tissue engineering (TE) approaches have emerged as potential solutions for OC lesions repair and regeneration 7. Since cartilage tissue is avascular and lacks of self-remodeling capacity, an important role is attributed to the subchondral bone in the regenerative process. Moreover, the articular cartilage is composed by four distinct zones (superficial, middle, deep and calcified cartilage zone) with different proportions in terms of ECM composition and cell maturation, being responsible for the collagen fibrils formation that will anchor the underlying subchondral bone 8. Thus, the recognition that the reconstruction of both articular cartilage and subchondral bone as two separate areas is overpassed, was an important achievement in the TE field. Still, a better knowledge of the mechanical properties, structure, and biological requirements of the two interconnected tissues, is required for the design of more complex scaffolding systems for OC applications. Bilayered scaffolds emerged as potential candidates for OC TE, mainly because they can mimic the structural complexity of the natural OC tissue 9. Different studies proposed homogeneous and heterogeneous bilayered scaffolds, using different biomaterials, scaffolding design, and fabrication methods, within in vitro and in vivo environments 10-12. For example, Li et al.

13

proposed biphasic scaffolds produced via a rotational combination of SF and bioactive ceramic, showing stratified

structural properties similar to the OC tissue, different pore morphologies depending on the cartilage or bone segment, as well as, chondrogenic and osteogenic differentiation potential in the two phases of the biphasic scaffolds. Using a different approach, Ding et al. 14 developed trilayered scaffolds combining SF/hydroxyapatite (HAp) and paraffin-sphere leaching with a thermal-induced phase separation (TIPS) technology, presenting porous and dense bony and interface layers, together with adequate properties for cell proliferation, infiltration and differentiation towards chondrocytes and osteoblasts at the chondral and subchondral layers. The “hierarchical” structure of OC scaffolds is another important concern, meaning that its features need to be comparable to that of the target tissue from the macroscopic down to the molecular level 15. For example, the nanoscale properties induced by the crosslinking process or HAp stoichiometry, positively influence cell adhesion and


3


Page 4 of 51

mineralization, whereas the micrometer-scale of the scaffolds porosity, affects cell migration, ECM formation, and at some extent provides the mechanical anisotropy of scaffolds. Several synthetic OC products were introduced in the market, including the Agili-CTM, MaioRegen® or the TruFit®

16.

However, due to the inconsistency of results proving the full efficacy of these structures in the repair/regeneration of OC tissue, only a few are clinically available 17-18. Natural polymers possess favorable biocompatibility and can easily promote cell adhesion 19. Among them, Bombyx mori silk fibroin (SF) has proved to exhibit tunable mechanical properties, being successfully used as scaffolding material for different TE applications

13, 20-21.

At the same time, calcium phosphates (CaPs) have shown remarkable osteoconductivity,

bioactivity and biocompatibility 22, which makes the combination of SF and CaPs attractive for bone and OC regeneration scaffolding 23-24. Among existing CaPs, β-tricalcium phosphate (β-TCP) has shown great osteconductivity and resorbability in vivo 24-26. Doping β-TCP with elements existing in bone, has shown enhanced mechanical properties of the scaffolds and directly affects the production and release of growth factors like vascular endothelial growth factor (VEGF), and bone morphogenetic protein-2 (BMP-2) involved on the osteogenic and angiogenic response of bone tissue 27-28. Up to date, three-dimensional (3D) networks, structural adaptability, and resilient mechanical properties of hydrogels make them attractive for several TE applications 29. Moreover, the possibility of structuring hydrogel-based systems by combining different processing technologies allowed the creation of robust matrices for high load-bearing applications

9, 30.

In this

context, we developed hydrogen peroxide (HRP)-crosslinked SF scaffolds for articular cartilage regeneration, showing that combining SF hydrogels with salt-leaching and freeze-drying techniques, can improve the stability and stiffness of the structures. In addition, the porosity required for extracellular matrix (ECM) formation and tissue ingrowth can be easily tuned 31.

Moreover, in the anti-parallel β-sheet form, these scaffolds presented higher stiffness and superior stability in response to

the enzymatic degradation. Thus, in this study we aim to combine these multifunctional tools in order to develop 3D bilayered and hierarchical scaffolds, aiming OC repair and regeneration. We hypothesize that through these strategies we will be able to create novel bilayered structures with controlled porosity at the macro-and micro-scale, and improved stability for shortto long-term OC implantation purposes. The spatially limited incorporation of bioresorbable β-TCP should create stratified models, affecting their mineralization and mechanical response as hard templates. Thus, a HRP-mediated approach was used to crosslink SF (HRP-SF) combined with pure and zinc and strontium (ZnSr)-doped β-TCP, through salt-leaching and freezedrying technologies. HRP-SF was used as articular cartilage-like layer, while a 80/20 (w/w) ratio of HRP-SF/ZnSr-doped βTCP (HRP-SF/dTCP) and HRP-SF/undoped β-TCP (HRP-SF/TCP) was used to produce the underlying subchondral bone layer. Taking into consideration the previous reported results obtained from the ion incorporation into monolayered SF-based scaffolds 28, where the combination of Zn and Sr positively influenced human adipose-derived stem cells (hASCs) growth


4

Page 5 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


and osteogenesis, in here we propose to evaluate the role of the same combination (Zn+Sr) for doping β-TCP related to physicochemical and biological behavior of the bilayered structures. Physicochemical evaluation and mechanical characterization were performed. The enzymatic degradation profile and bioactivity of the scaffolds were evaluated by soaking in protease XIV and simulated body fluid (SBF) solutions, respectively, up to 30 days. The in vitro cell viability, adhesion and proliferation were also assessed by co-culturing primary human osteoblasts (hOBs) and human articular chondrocytes (hACs) into the bilayered structures up to 14 days.

2. Experimental Section 2.1. Materials and reagents Bombyx mori cocoons were provided by the Portuguese Association of Parents and Friends of Mentally Disabled Citizens (APPACDM, Castelo Branco, Portugal). Silicone tubing was acquired from Deltalab (Barcelona, Spain). Granular sodium chloride low in endotoxins (EMPROVE®, VWR BDH Prolabo, Briare, France) was sieved at 500-1000 µm range (Analytic Sieve Shaker; AS 200 Digit, Retsch, Germany). All reagents were purchased from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise described. 2.2. Preparation of the bilayered scaffolds Purified aqueous SF solution were extracted from Bombyx mori cocoons and obtained at high concentration (16 wt.%), following a previously reported procedure

21.

Powders of β-TCP, pure and doped with 10 mol.% of Zn and Sr, were

synthesized by aqueous precipitation, and characterized as described in a previous work 28. The cartilage-like layer of the scaffolds was produced with the HRP-crosslinked SF (HRP-SF) solution at 1/0.26‰/1.45‰ (SF/HRP/H2O2), while the subchondral bone-like layer was prepared with 80/20 (w/w) ratio HRP-SF/ZnSr-β-TCP, and HRP-SF/β-TCP with no ion incorporation, for comparison purposes. Firstly, the subchondral bone-like layers were produced by mixing ZnSr-β-TCP or β-TCP powders with the HRP-SF solution and transferred into cylindrical shaped silicone molds (9 mm inner ), followed by adding 2 g of granular sodium chloride (particle size 500-1,000 µm). After complete gelation at 37ºC in the oven, the molds were soaked in distilled water and the salt was subsequently extracted (porogen) for 72 hours. Finally, a biopsy punch (8 mm inner diameter; Smith & Nephew, Portugal) was used to remove the subchondral bone-like layers’ scaffolds from the molds, and cut into pieces. Secondly, the subchondral bone-like layers were placed in the bottom of new silicon molds (8 mm inner ) and the HRP-SF solution was added on the top of these scaffolds in order to produce hierarchical bilayered scaffolds, followed by adding granular sodium chloride particles. After complete gelation and salt-leaching process, the bilayered scaffolds were frozen at -80ºC overnight, and freeze-dried (Telstar Cryodos-80, Barcelona, Spain) during 7 days. Monolayers of HRP-SF, HRP-SF/ZnSr-β-TCP (HRP-SF/dTCP), and HRP-SF/β-TCP (HRP-SF/TCP) scaffolds, were


5


Page 6 of 51

prepared as controls. The bilayered scaffolds are abbreviated as: BdTCP for bilayered HRP-SF|HRP-SF/dTCP scaffolds, and BTCP for bilayered HRP-SF|HRP-SF/TCP scaffolds (Scheme 1a). 2.3. Physicochemical characterization 2.3.1. Scanning electron microscopy and energy dispersive spectroscopy analysis The morphology of the scaffolds was analyzed by scanning electron microscopy (SEM) using a JEOL JSM-6010LV (Tokyo, Japan). Before the analysis, all samples were sputter-coated with gold (Leica EM ACE600 coater; Leica Microsystems, Wien, Austria). The elemental composition was performed by energy dispersive spectroscopy (EDS; Pegasus X4M) coupled to the SEM. Three independent regions were selected in the BdTCP and BTCP scaffolds, corresponding to the HRP-SF layer, interface, and HRP-SF/dTCP or HRP-SF/TCP layer. 2.3.2. Micro-computed tomography The quantitative and qualitative evaluation of the BdTCP and BTCP scaffolds’ microstructure were performed using a highresolution X-ray microtomography system (Skyscan 1272; Skyscan, Kontich, Belgium). The scanning of the scaffolds was conducted using a pixel size of 3.5 µm and an X-ray source fixed at 50 keV and 200 µA. The scaffolds diameter and height were 8 mm (HRP-SF layer: 3 mm height; HRP-SF/dTCP and HRP-SF/TCP layer: 5 mm height). The X-ray projections were reconstructed to obtain the two-dimensional (2D) images. The 2D images in each data set (HRP-SF layer, HRP-SF/dTCP layer and HRP-SF/TCP layer) were binarized automatically with a global threshold using the manufacturer's software (CT Analyzer v1.17, SkyScan, Kontich, Belgium) to quantify the mean of porosity, pore size, wall thickness and interconnectivity. The porosity and TCP content distribution profiles were also determined on the 2D images. 2D and 3D qualitative visualization of the different phases at the BdTCP and BTCP scaffolds were obtained with Data Viewer (v1.7.1.0) and CTVox (v3.3.0.r1412) (SkyScan, Kontich, Belgium), respectively. The quantitative and qualitative evaluation of scaffolds was performed using three samples per scaffold. 2.3.3. X-ray diffraction The qualitative analysis of crystalline phases presented on the HRP-SF, HRP-SF/dTCP and HRP-SF/TCP scaffolds was determined by X-ray diffraction (XRD) using a high-resolution Bragg–Brentano diffractometer (Bruker D8 Advance DaVinci, Karlsruhe, Germany) equipped with CuKα radiation (λ = 1.542 Å), produced at 40 kV and 40 mA. All data were collected in the 2 range of 10–60º with a step size of 0.02º and 1 second per step. The analysis of each condition was repeated three times independently. 2.3.4. Mechanical compressive tests The compressive tests of the BdTCP and BTCP scaffolds were performed in wet state using a Universal Testing Machine


6

Page 7 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(Instron 4505, Norwood, MA, USA). The scaffolds diameter and height were 8 mm (HRP-SF layer: 3 mm height; HRPSF/dTCP and HRP-SF/TCP layer: 5 mm height). Before analysis, the samples were immersed in phosphate-buffered saline (PBS, pH 7.4) solution overnight (until equilibrium was reached). The cross-head speed was fixed at 2 mm/min and tests were proceeded until a 60% deformation. The elastic modulus (E) was determined through the slope of the initial linear section of the stress-strain curve obtained for each sample. Five samples of each group were used to determine the average compressive modulus and compressive strength. HRP-SF, HRP-SF/dTCP and HRP-SF/TCP scaffolds were used as controls (8 mm diameter and 4 mm height). 2.3.5. Dynamic mechanical analysis The dynamic mechanical behavior (DMA) of the BdTCP and BTCP scaffolds was determined in a mechanical analyzer (TRITEC8000B; Triton Technology, Lincolnshire, UK) at compressive mode, using the same samples sizes as the ones for the compressive tests. The measurements were performed in hydrated state at 37ºC, by immersing the samples in PBS solution, overnight, until reaching equilibrium at 37ºC. The DMA spectra were acquired in a frequency scan from 0.1 to 10 Hz, used to obtain both storage modulus (E’) and loss factor (tan δ). During experiments, a constant strain amplitude of 50 µm was used. Before the analysis, each sample was subjected to a small pre-load, to ensure the contact between all the surface and the compression plates, with equivalent distance for all tested samples. Five samples of each group were tested and HRP-SF, HRP-SF/dTCP and HRP-SF/TCP scaffolds were used as controls (8 mm diameter and 4 mm height). 2.4. Degradation profile The stability of the BdTCP and BTCP scaffolds was determined by enzymatic degradation test. Protease XIV (Streptomyces griseus, 3.5 U/mg) was dissolved in PBS solution at 2 U/mL and 0.0035 U/mL

32.

Samples were also incubated in PBS

solution to be used as controls. The initial wet weight of each sample (n=3) was measured after hydration in PBS solution for 3 hours at 37ºC, and then each scaffold was immersed in 5 mL of protease solution or fresh PBS solution. The study was conducted at 37ºC for a time period ranging from 6 hours to 30 days. Fresh enzyme solutions were used every 24 hours. At the end of each time point, the wet weight was measured after washing the samples with distilled water. Before weighing, samples were gently blotted with a filter paper to remove the excess liquid. The weight loss was determined as shown in the following equation:

Weight loss (%) =

𝑚𝑖 ― 𝑚𝑓 𝑚𝑖

× 100

where mi and mf are the wet weight of respectively, initial and degraded samples at different time point. HRP-SF, HRPSF/dTCP and HRP-SF/TCP scaffolds were used as controls.


7


Page 8 of 51

2.5. In vitro bioactivity assay The in vitro bioactivity of the BdTCP and BTCP scaffolds was evaluated by immersion in SBF solution for 15 and 30 days. Samples were immersed in SBF solution containing ion concentrations (Na+ 142.0, K+ 5.0, Ca2+ 2.5, Mg2+ 1.5, Cl- 148.8, HPO4- 1.0, HCO32+ 4.2, SO42- 0.5 mM.L-1, pH=7.4) similar to those of human blood plasma, at 37ºC in continuous shaking (60 rpm) 33. After each time point,the samples were rinsed in distilled water and allowed to dry at 37ºC for 3 days, followed by SEM and EDS analysis. Three samples of each group were tested and HRP-SF, HRP-SF/dTCP and HRP-SF/TCP scaffolds were used as controls. 2.6. In vitro cell studies 2.6.1. Culture of hOBs and hACs Primary human osteoblast (hOBs) obtained from femoral bone tissue, and human articular chondrocytes (hACs) obtained from knee articular cartilage (Innoprot, Derio, Spain), were used in this study. Cells were expanded in culture medium consisting of DMEM/F12 (Dulbecco's modified Eagle's medium/F12) GlutaMAX Supplement (Gibco®, Life Technologies, Carlsbad, CA, USA) supplemented with 10% (v/v) fetal bovine serum (FBS; Life Technologies, Carlsbad, CA, USA), 1% (v/v) antibiotic/antimycotic (100 units/mL penicillin and 100 mg/mL streptomycin, Life Technologies, Carlsbad, CA, USA), 100 nM Dexamethasone, 1% (v/v) MEM-NEAA (Gibco®, Life Technologies, Carlsbad, CA, USA) and 150 µg/mL Ascorbic acid. Cells were cultured until confluence at 37ºC, 5% CO2 incubator, and culture medium was changed every 2-3 days. 2.6.2. Seeding and co-culture of hOBs-hACs on the bilayered scaffolds For the co-culture system, BdTCP and BTCP scaffolds of 8 mm diameter and height (HRP-SF layer: 3 mm height; HRPSF/dTCP and HRP-SF/TCP layer: 5mm height) were used and sterilized by β-radiation (IONMED sterilization SA, Tarancón, Cuenca, Spain). Before the cell seeding, all scaffolds were hydrated in DMEM/F12, Glutamax supplemented with 1% (v/v) antibiotic/antimycotic solution, in the CO2 incubator overnight. The hydrated scaffolds were cut longitudinally with a blade and transferred to 24-well culture plates (Corning Incorporated, Life Sciences, Durham, NC, USA). Human OBs of passage 5 were detached from the flasks and a 10 µL cell suspension was used for seeding onto the surface of the HRPSF/dTCP and HRP-SF/TCP layers, at a density of 3 x 105 cells/scaffold (P6). The constructs were mantained in the CO2 incubator for 3 hours and then completed with 2 mL of DMEM/F12, Glutamax supplemented with 10% (v/v) FBS, 1% (v/v) antibiotic/antimycotic solution, 50 mM Dexamethasone, 1% (v/v) MEM-NEAA and 150 µg/mL Ascorbic acid. After 24 hours, the hACs were seeded (P6) onto the surface of the HRP-SF layers, following the same procedure described for seeding the hOBs. HRP-SF scaffolds (8 mm diameter and 3 mm height), HRP-SF/dTCP and HRP-SF/TCP scaffolds (8 mm diameter and 5 mm height) were seeded with hACs and hOBs, respectively, and used as monoculture controls (Scheme 1b). The cell-


8

Page 9 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


seeded constructs used for each characterization technique will be described below. Samples were collected after culturing for 1, 7 and 14 days, and culture medium changed every 2-3 days. 2.6.3. Metabolic activity of hOBs-hACs on the bilayered scaffolds The metabolic activity of cells in the constructs was screened by AlamarBlue® (BioRad, Hercules, CA, USA) assay, following the manufacturer’s instructions. After 3 hours of reaction with cells in a CO2 incubator at 37ºC, the Alamar blue solution was transferred into a 96-well culture plate (Greiner Bio-one, Frickenhausen, Germany) in triplicate (100 µL by well). A microplate reader (Synergy HT, BioTek, Instruments, USA) was used for fluorescence measurements at 530/590 nm. Three independent experiments were performed using three samples per group at each time-point. Scaffolds without cells were used as control. 2.6.4. Adhesion of hOBs-hACs on the bilayered scaffolds Cell adhesion and distribution were observed in the constructs by SEM (Model XL 30, Philips, Eindhoven, The Netherlands). After each culture time, the constructs were rinsed in PBS solution and fixed for 1 hour at 4ºC with a 2.5% (v/v) glutaraldehyde solution. A series of ethanol concentrations (30, 50, 70, 90 and 10% v/v) were used for samples dehydration, twice each concentration during 15 minutes, and left overnight at room temperature (RT) to dry. Prior analysis, all scaffolds were sputter-coated with gold-palladium (60:40; Polaron SC 7640, Watford, UK). 2.6.5. Cell proliferation assessment Cell proliferation was assessed using a fluorimetric double stranded DNA (dsDNA) quantification kit (Quant-IT PicoGreen dsDNA kit; Life Technologies, Carlsbad, CA, USA), according to the manufacturer’s instructions and following a previously reported procedure 31. The fluorescence intensity of samples was measured using a microplate reader at 485/528 nm and DNA content determined using a standard curve prepared with concentrations ranging from 0 to 2 µg/ml. Cell number was determined using cell standard curves prepared for the co-culture and monoculture systems, ranging from 3 x 105 to 1.6 x 106 cells/mL and 1.5 x 105 to 8 x 105 cells/mL, respectively. Three independent experiments were performed using three samples per group at each time-point. Scaffolds without cells were used as control. 2.6.6. Alkaline phosphatase activity quantification The lysates were used for DNA assay and alkaline phosphatase (ALP) activity quantification, following a previously reported procedure 20. Samples absorbance was recorded in a microplate reader at 405 nm and ALP concentrations determined based on a standard curve prepared with p-nitrophenol (pNP) solution with values ranging from 0 to 0.2 µmol/mL. Cell number was used to normalize the obtained ALP values of the same samples. Three independent experiments were performed using three samples per group at each time-point. Scaffolds without cells were used as control.


9


Page 10 of 51

2.6.7. Glycosaminoglycans quantification The glycosaminoglycans (GAGs) quantification was performed using a dimethylmethylene blue (DMB) assay, following a previously reported procedure 31. The optical density (OD) of samples was measured using a microplate reader at 530 nm absorbance and GAGs concentrations determined based on a standard curve prepared with chondroitin sulfate (Bioiberica, SA, Barcelona, Spain) solution (50 μg/mL in distilled water) with concentrations ranging from 0 to 35 µg/m. Three independent experiments were performed using three samples per group at each time-point. Scaffolds without cells were used as control. 2.6.8. Histology and immunofluorescence staining Samples were collected after each time-point and processed for histology. Constructs were transferred to histological cassettes and fixed with 10% (v/v) formalin solution. After paraffin-embedding, samples were sectioned with 20 µm thick in a microtome (Spencer 820, American Optical Company, NY, USA). Standard Hematoxylin & Eosin (Thermo Scientific, Waltham, MA, USA) staining was performed to evaluate cell distribution and ECM formation within the constructs. The collagen present in the ECM was detected by Sirius red/Fast green collagen staining kit (Chondrex, Redmond, WA, USA), where sirius red dye specifically binds to collagen and the non-collagenous proteins were stained by Fast green dye. Briefly, a 0.1% (v/v) Sirius red and 0.1% (v/v) Fast green mixture solution, saturated in picric acid, was used for sections stained. Safranin-O (0.1% v/v; Honeywell Fluka, Morris Plains, NJ, USA) staining was used to detect the distribution of sulfated GAGs. For counterstaining, sections were immersed with Gill-2 hematoxylin (Thermo Scientific, Waltham, MA, USA) and Fast green (0.02% v/v; Honeywell Fluka, Morris Plains, NJ, USA). Staining with 2% (w/v) Alizarin red solution (Merck, Germany) prepared in ultrapure water, was performed to assess the matrix mineralization. For protein expression detection, the sections were firstly submitted to antigen retrieval and permeabilization following a procedure reported elsewhere

31.

Immunolabeling was performed using a rabbit anti-human polyclonal antibody against

osteopontin (OPN; Abcam, Cambridge, UK; dilution 1:25), a mouse anti-human monoclonal antibody against aggrecan (ACAN; clone BC-3, Thermo Scientific, Waltham, MA, USA; dilution 1:5), a mouse anti-human monoclonal antibody against osteocalcin (OCN; clone OC4-30, Abcam, Cambridge, UK; dilution 1:25), and a rabbit anti-human polyclonal antibody against collagen X (Col X; Abcam, Cambridge, UK; dilution 1:100) as primary antibodies, incubated overnight at 4ºC. The following incubation with the respective secondary fluorochrome-conjugated antibodies, anti-rabbit/mouse IgG (Invitrogen, Life Technologies, California, USA; dilution 1:100), as conducted for 2 hours in the dark at RT. A 4,6Diamidino-2-phenyindole, dilactate (DAPI; Biotium, CA, USA; dilution 1:500) solution was used for nuclei staining, incubated for 15 minutes. Sections incubated only with the secondary fluorochrome-conjugated antibodies were used as negative controls (Figure S1).


10

Page 11 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


A transmitted and reflected light microscope (Axio Imager Z1 m; Zeiss, Jena, Germany) was used for histological and immunofluorescent (OPN in green: ex/em 488/517; ACAN in red: ex/em 594/618; DAPI in blue: ex/em 358/461) sections analysis. Images were obtained using the Zen microscope software (Zeiss, Jena, Germany), connected to the digital cameras AxioCam MRc5 and MR3 (Zeiss, Jena, Germany). 2.6.9. RNA isolation and real-time quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) The total mRNA from the BdTCP and BTCP constructs was extracted using the Direct-zolTM RNA MiniPrep kit (Zymo Research, Irvine, CA, USA), following the manufacturer’s instructions. Briefly, after 1 and 14 days of culture, the constructs were washed with PBS solution, cut through their different layers (HRP-SF, interface, HRP-SF/dTCP or HRP-SF/TCP) and immersed in 200 µL TRI Reagent for storing at -80ºC until further use. Samples were thawed at RT and sonicated using an ultrasonic processor (Sonics Materials™ VCX-130PB Ultrasonic Processor, Newtown, USA) to ensure the complete lysing and removal of cells from the constructs. The total RNA pellets were reconstructed in RNase free water (Gibco®, Life Technologies, Carlsbad, CA, USA). For RNA quantification determination, a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA) was used. Complementary DNA (cDNA) synthesis was performed according to the protocol from the qScriptTM cDNA synthesis Kit (Quanta Biosciences, Gaithersberg, MD, USA) using a MiniOpticon real-time PCR detection system (BioRad, Hercules, CA, USA), where 100 ng of total RNA were used to obtain singlestranded cDNA with qScript Reverse Transcriptase (RT). The cDNA was further used as template for the amplification of osteogenic- (OCN and OPN) and chondrogenic-related (ACAN, Sox-9 and Col X) genes (Table 1) using the PerfeCta SYBR Green FastMix kit (Quanta Biosciences, Gaithersberg, MD, USA) according to the manufacturer’s instructions. Forty-five cycles of denaturation (95 ºC, 10 s), annealing (temperature specific for each gene, 25 s) and extension (72 ºC, 30 s) were carried out in a Mastercycler ep realplex real-time PCR system (Eppendorf, Hamburg, Germany). The transcripts expression data of each sample were normalized to the housekeeping gene glyceraldehyde-3-phosphate-dehygrogenase (GAPDH) of that sample, for each tested time point. The relative gene expression was calculated according to the Livak (2-ΔΔCt) method, considering the Day 1 condition as the calibrator. A parallel study was conducted to evaluate the hOB and hACs cell growth separately in the co-culture medium for 1, 7 and 14 days. Relative expression of osteogenic- and chondrogenic-related genes was determined (Figure S2).


11


Page 12 of 51

Table 1. Primers list for the osteogenic- and chondrogenic-related markers. Sequences Gene Forward (5’-3’)

Reverse (5’-3’)

GADPH

TGCACCACCAACTGCTTAGC

GGCATGGACTGTGGTCATGAG

OCN

CTGGAGAGGAGCAGAACTGG

GGCAGCGAGGTAGTGAAGAG

OPN

CAGACCTGACATCCAGTACCC

GGTCATCCAGCTGACTCGTT

Sox-9

TACGACTACACCGACCACCA

TTAGGATCATCTCGGCCATC

ACAN

TGAGTCCTCAAGCCTCCTGT

TGGTCTGCAGCAGTTGATTC

Col X

CCAGGTCTCGATGGTCCTAA

GTCCTCCAACTCCAGGATCA

2.7. Statistical analysis All the numerical results are presented as mean ± standard deviation (SD). The GraphPad Prism 5.0 software (GraphPad Software, La Jolla, CA, USA) was used for performing statistical analysis. First, a Shapiro-Wilk test was applied to verify data normality. The results obtained from static mechanical properties were evaluated by means of a Mann-Whitney test. A Kruskal-Wallis test and Dunn's multiple comparison test were used to ascertain the differences in the experimental results from metabolic activity, proliferation, ALP activity and GAG’s content. The results from real-time PCR analysis were analyzed by means of a Mann-Whitney test. The significance level was set to *p < 0.05, **p < 0.01, ***p < 0.001.


12

Page 13 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Scheme 1. Schematics of the experimental setup adopted in this study. (a) Methodology used for preparing the BdTCP and BTCP scaffolds, creating HRP-SF scaffolds as cartilage-like layers in combination with HRP-SF/dTCP and HRP-SF/TCP scaffolds in an 80/20 (w/w) ratio as subchondral bone-like layers. (b) Temporal approach followed for co-culturing human osteoblasts (hOBs) and human articular chondrocytes (hACs) on the corresponding HRP-SF/dTCP, HRP-SF/TCP and HRPSF layers of the BdTCP and BTCP scaffolds. Monoculture system of hOBs on the monolayered HRP-SF/dTCP and HRPSF/TCP scaffolds, and hACs on the monolayered HRP-SF scaffold used as controls. Scaffolds were prepared with 8 mm diameter and height (HRP-SF layer: 3 mm in height; HRP-SF/dTCP and HRP-SF/TCP layers: 5 mm in height).

3. Results 3.1. Microstructure, elemental composition and TCP distribution into the bilayered scaffolds Figure 1 shows the microstructure of BdTCP and BTCP scaffolds observed by SEM and the chemical elements analyzed using EDS. It can be observed that the bilayered scaffolds presented a macro- and micro-porous structure on both HRP-SF (Figure 1a and d), and HRP-SF/dTCP (Figure 1c) or HRP-SF/TCP (Figure 1f) layers, presenting macro-pores larger than 500 μm and micro-pores that reach 10 μm. The scaffold layers were well integrated by continuous interface regions of ~500 μm thickness (Figure S3). From EDS spectra, it was possible to detect calcium (Ca) and phosphorous (P) ions in the subchondral bone-like layers (HRP-SF/dTCP and HRP-SF/TCP) (Figure 1c, f) and interface regions (Figure 1b and e), as expected. The


13


Page 14 of 51

intensity of the Ca and P peaks was higher on the HRP-SF/dTCP and HRP-SF/TCP layers than in the interface regions, reaching Ca/P ratio values between 1.53 and 2.31 (Figure 1b, c, e and f). It was also detected the presence of Zn and Sr on the doped scaffolds, namely BdTCP interface (Figure 1b) and HRP-SF/dTCP layer (Figure 1c). The qualitative and quantitative analysis of the BdTCP and BTCP scaffolds architecture were assessed by micro-CT. Table 2, shows that the bilayered scaffolds presented similar porosity and pore wall thickness, with high interconnectivity of 79.7 ± 5.9% on BdTCP scaffolds, and 75.2 ± 3.8% on BTCP scaffolds. As the bilayered scaffolds porosity decreased, from the HRP-SF layers (89.1 ± 1.4% on BdTCP; 89.1 ± 0.6% on BTCP) to the HRP-SF/dTCP (31.4 ± 8.4%) and HRP-SF/TCP (21.6 ± 8.8%) layers, including the respective interfaces (76.4 ± 2.8% on BdTCP; 71.6 ± 14.6% on BTCP), the mean pore wall thickness tends to increase on both BdTCP (HRP-SF layer: 22.6 ± 2.0 µm; Interface: 47.8 ± 5.5 µm; HRP-SF/dTCP layer: 50.7 ± 4.4 µm) and BTCP (HRP-SF layer: 21.5 ± 2.5 µm; Interface: 51.9 ± 2.3 µm; HRP-SF/TCP layer: 59.2 ± 8.4 µm) scaffolds. From the 2D images, it was possible to observe the porous structure in each layer of the BdTCP (Figure 2a and b) and BTCP scaffolds (Figure 2d and e). Moreover, the TCP component (ZnSr-β-TCP and β-TCP) was retained only in the hybrid layers (Figure 2b and e), as confirmed by the blue domain present in the 3D reconstructions of the BdTCP (Figure 2c) and BTCP (Figure 2f) scaffolds. The porosity distribution profile was homogeneous in each scaffold layer (Figure 2g); however, a substantial increase of porosity was observed from the interface region until the HRP-SF layers. The TCP component was evenly distributed in the HRP-SF/dTCP and HRP-SF/TCP layers, gradually decreasing at the interface domain and being completely absent at the HRP-SF layers (Figure 2h).

Figure 1. SEM micrographs (scale bar: 100 μm) and EDS elemental analysis with the detected ion elements and Ca/P ratios determined at the different regions of the BdTCP and BTCP scaffolds. (a, d) HRP-SF layers, (b, e) interface, (c) HRPSF/dTCP layer, and (f) HRP-SF/TCP layer.


14

Page 15 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 2. Microstructure of the BdTCP and BTCP scaffolds analyzed by micro-CT. Mean porosity (%)

BdTCP

BTCP

Mean wall thickness (µm)

HRP-SF

89.1 ± 1.4

22.6 ± 2.0

Interface

76.4 ± 2.8

47.8 ± 5.5

HRP-SF/dTCP

31.4 ± 8.4

50.7 ± 4.4

HRP-SF

89.1 ± 0.6

21.5 ± 2.5

Interface

71.7 ± 14.6

51.9 ± 2.3

HRP-SF/TCP

21.6 ± 8.8

59.2 ± 8.4

Mean interconnectivity (%)

79.7 ± 5.9

75. 2 ± 3.8

Figure 2. Micro-CT analysis of the BdTCP and BTCP scaffolds. 2D images of the (a, d) HRP-SF layers, (b) HRP-SF/dTCP layer and (e) HRP-SF/TCP layer (scale bar: 2 mm). 3D reconstructions of the (c) BdTCP scaffolds and (f) BTCP scaffolds showing the HRP-SF matrix in brown and the HRP-SF/dTCP or HRP-SF/TCP matrices in blue (scale bar: 2 mm). Change of (g) 2D porosity distribution across the vertical axis of the scaffolds and (h) TCP (ZnSr-β-TCP and β-TCP) component distribution from the subchondral bone-like layers across the vertical axis of the scaffolds (i.e. HRP-SF/dTCP and HRPSF/TCP, respectively) to the cartilage-like layers (HRP-SF).


15


Page 16 of 51

3.2. Crystalline structure and mechanical properties of the scaffolds XRD patterns of the HRP-SF, HRP-SF/dTCP and HRP-SF/TCP scaffolds are displayed in Figure 3a. All the scaffolds showed the typical crystalline peaks of β-sheet structure (silk-II conformation), being more pronounced for HRP-SF scaffolds with main peaks located at 20.5º and 24.3º

21, 34.

Additionally, the HRP-SF/dTCP and HRP-SF/TCP scaffolds showed the

characteristic crystalline phases of β-TCP (standard ICDD PDF 04-014-2292) and trace amounts of β-calcium pyrophosphate (β-CPP) (standard ICDD PDF 04-009-3876). A small peak shift at 2θ ∼ 28º toward higher 2θ, namely in the ZnSr-β-TCP powders, is also observed in these scaffolds due to the incorporation of ion-dopants in the crystal structure (β-sheet) of SF (zoomed area in Figure 3a). The observed variations can be understood considering the ion radii of the Zn (0.74 Å) and Sr (1.12 Å) ions, in comparison to Ca (0.99 Å), thus affecting the unit of the lattice parameters of the β-sheet structure 35. As shown in Figure 3b, the wet compressive modulus of the BdTCP (226.56 ± 60.34 kPa) and BTCP (189.18 ± 90.80 kPa) scaffolds was superior than that obtained for the corresponding monolayered scaffolds (HRP-SF: 56.46 ± 15.34 kPa; HRPSF/dTCP: 156.88 ± 92.36 kPa; HRP-SF/TCP: 147.5 ± 82.07 kPa). A similar pattern was obtained when examined the compressive strength of HRP-SF (27.76 ± 3.68 kPa), HRP-SF/dTCP (73.41 ± 33.71 kPa), HRP-SF/TCP (68.97 ± 36.14 kPa), BdTCP (124.84 ± 23.42 kPa) and BTCP (91.99 ± 6.08 kPa) scaffolds (Figure 3c). The HRP-SF scaffolds presented a significantly lower compressive modulus and compressive strength, as compared to the BdTCP (**p < 0.01 and *p < 0.05, respectively) and BTCP (*p < 0.05) scaffolds. Moreover, a significantly higher compressive strength was presented by the BdTCP scaffolds (*p < 0.05), as compared to the pure BTCP. From the dynamic viscoelastic properties assessed by DMA (Figure 3d and e), it was found that the storage modulus (E’) of the bilayered and monolayered control scaffolds increased at lower rates, with increment of the frequencies (from 0.1 to 10 Hz), ranging from 403.56 ± 111.62 to 593.42 ± 206.90 kPa on BdTCP scaffolds, 262.72 ± 59.92 to 347.68 ± 93.37 kPa on BTCP scaffolds, 180.43 ± 52.84 to 238.40 ± 94.48 kPa on HRP-SF scaffolds, 425.58 ± 329.18 to 617.24 ± 505.05 kPa on HRP-SF/dTCP scaffolds and 267.63 ± 90.69 to 371.22 ±153.76 kPa on HRP-SF/TCP scaffolds. In the frequency range tested, the E’ values of BdTCP and BTCP scaffolds were similar to those of HRP-SF/dTCP and HRP-SF/TCP, respectively, and higher than those of HRP-SF scaffolds, indicating that the presence of TCP instigated the reinforcement of the structure, increasing the mechanical strength. The loss factor (tan δ) obtained for the bilayered and monolayered control scaffolds were constant when the frequency increased from 0.1 to 10 Hz. All groups of scaffolds presented similar and high loss factor values for the frequencies tested.


16

Page 17 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. Crystalline structure and mechanical characterization of the BdTCP and BTCP scaffolds: (a) XRD patterns of the HRP-SF, HRP-SF/dTCP and HRP-SF/TCP scaffolds, and zoomed area of graph in a 2θ range from 27 to 36, showing the shift toward higher 2θ in the HRP-SF/dTCP scaffolds. (b) Compressive modulus and (c) compressive strength of the bilayered scaffolds (BdTCP and BTCP) and controls (HRP-SF, HRP-SF/dTCP and HRP-SF/TCP), measured in the hydrated state. (d) Storage modulus (E’) and (e) loss factor (tan 𝜹) of the bilayered scaffolds (BdTCP and BTCP) and controls (HRP-SF, HRPSF/dTCP and HRP-SF/TCP) obtained by DMA, tested at physiological conditions (pH 7.4 and 37ºC). 3.3. Degradation properties and bioactivity evaluation of the scaffolds The enzymatic degradation analysis of the bilayered scaffolds, and monolayered scaffolds used as control, was performed using protease XIV (Figure 4a) 36. As expected, the degradation rate of all tested groups was superior in protease XIV at the highest tested concentration (2 U/mL), as compared to the protease XIV concentration of 0.0035 U/mL and to the control condition (PBS solution), as SF is prone to proteolytic degradation. HRP-SF scaffolds presented higher weight loss than the bilayered scaffolds, and remaining monolayered control groups, in the presence of protease XIV at 2 U/mL. Nevertheless, after 3 days of testing, all the monolayered control scaffolds were completely degraded, while the BdTCP and BTCP scaffolds reached 100% weight loss after 14 and 21 days, respectively. The weight loss of samples in contact with 0.0035 U/mL of protease XIV was stable over the first 7 days for all tested groups. However, a substantial decrease in weight loss was


17


Page 18 of 51

observed after 21 and 30 days, reaching a weight loss of ~70% on BdTCP scaffolds, ~80% on BTCP scaffolds and HRP-SF scaffolds, ~90% on HRP-SF/dTCP scaffolds, and 100% on HRP-SF/TCP scaffolds. In the control conditions, all scaffolds immersed in PBS solution maintained their original weight for 30 days. Figure 4b, shows the morphology of the bilayered and monolayered control scaffolds after immersion in SBF for 15 days. It was observed that the HRP-SF/dTCP (Figure 4bii and vi) and HRP-SF/TCP (Figure 4biv and vii) samples induced the formation of cauliflower-like apatite crystals, while there were no apatite crystals formed on the HRP-SF layers (Figure 4bi and iii) or scaffolds (Figure 4v). The formation of a mineral layer at the HRP-SF/dTCP and HRP-SF/TCP layers, with similar morphology to that of HAp, was also confirmed by EDS analysis showing Ca/P ratio values between 1.93 and 2.43 (Figure 4bii, iv, vi and vii).

Figure 4. Degradation profile and in vitro bioactivity evaluation of the BdTCP and BTCP scaffolds. (a) Enzymatic degradation of the bilayered scaffolds (BdTCP and BTCP) and controls (HRP-SF, HRP-SF/dTCP and HRP-SF/TCP), measured for a period of 30 days (% weight). Samples immersed in PBS were used as controls. (b) SEM micrographs (scale bar: 10 μm) and EDS elemental analysis with the detected ion elements and Ca/P ratios determined at the (i, iii) HRP-SF layers, (ii) HRP-SF/dTCP layer and (iv) HRP-SF/TCP layer of the BdTCP and BTCP scaffolds, after 15 days of immersion in SBF solution. (v) HRP-SF scaffolds, (vi) HRP-SF/dTCP scaffolds and (vii) HRP-SF/TCP scaffolds were also evaluated. 3.4. In vitro characterization of the bilayered scaffolds 3.4.1. Cell adhesion and metabolic activity Alamar blue assay was performed in order to evaluate the metabolic activity of cells up to 14 days in co-culture and monoculture system (Figure 5). It was observed that the metabolic activity of cells on both co-culture and monoculture systems, significantly increased from day 1 to day 7 (*p < 0.05 on HRP-SF/TCP; **p < 0.01 on BdTCP; ***p < 0.001 on HRP-SF and BTCP), from day 1 to day 14 (***p < 0.001), and from 7 to day 14 (*p < 0.05, except for HRP-SF ***p < 0.001). The BdTCP and BTCP scaffolds co-cultured with hOBs and hACs, did not present significant differences in metabolic


18

Page 19 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


activity at the same culture periods. The same behavior was observed between the HRP-SF/dTCP and HRP-SF/TCP scaffolds monocultured with hOBs. From SEM analysis (Figure 6), it was possible to observe from day 1 that the hACs and hOBs were able to adhere on the micro-porous structures, showing a high degree of spreading and extended lamellipodia over the scaffolds surface, both on the bilayered and monolayered scaffolds. The macro-porosity also allowed cells to penetrate into the scaffolds.

Figure 5. Metabolic activity of the hOBs and hACs co-cultured in the BdTCP and BTCP scaffolds for 1, 7 and 14 days, determined by Alamar blue assay. Monocultures of hACs in the HRP-SF scaffolds and hOBs in the HRP-SF/dTCP and HRPSF/TCP scaffolds were used as controls.


19


Page 20 of 51

Figure 6. SEM micrographs of the co-cultured hACs and hOBs for 1, 7 and 14 days, on the corresponding HRP-SF layer, HRP-SF/dTCP or HRP-SF/TCP layers of the BdTCP and BTCP scaffolds, respectively (scale bar: 100 μm). Monocultures of hACs in the monolayered HRP-SF scaffolds, and hOBs in the monolayered HRP-SF/dTCP and HRP-SF/TCP scaffolds were observed as controls. The yellow arrows indicate the cells inside the macro-pores. 3.4.2. Evaluation of osteogenesis and chondrogenesis Cell proliferation was quantified using the cell content based on dsDNA quantification (Figure 7a). A significant increase of the cell number was observed up to 14 days of culture on both co-culture and monoculture systems. The bilayered scaffolds co-cultured with hOBs and hACs, presented a significant increase of the cell number from day 1 to day 14 (***p < 0.001), and from day 1 to day 7 (***p < 0.001 on BTCP). Regarding the HRP-SF constructs, a significant increase of the cell number was detected from day 1 to day 14 (***p < 0.001), and from day 7 to day 14 (*p < 0.05). On the monoculture system with hOBs, a significant increase of cell proliferation was observed from day 1 to day 14 on both HRP-SF/dTCP (***p < 0.001)


20

Page 21 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


and HRP-SF/TCP (*p < 0.05) constructs, and from day 1 to day 7 on the HRP-SF/dTCP constructs (*p < 0.05). No significant differences were observed when compared the hOBs proliferation on the HRP-SF/dTCP and HRP-SF/TCP scaffolds, at the same culture periods. The co-culture groups presented the same trend. The ALP activity of the mono- and co-cultured cells was normalized by the corresponding number of cells in the same scaffolds (Figure 7b). In co-culture conditions, the ALP activity significantly increased on both BdTCP (***p < 0.001 from day 1 to day 14; **p < 0.01 from day 7 to day 14) and BTCP (**p < 0.01 from day 1 to day 7; ***p < 0.001 from day 1 to day 14; *p < 0.05 from day 7 to day 14) constructs. Nevertheless, no significant differences were observed when compared the ALP activity in the bilayered constructs at the same culture periods. It was observed that the ALP activity of hACs monocultured in the HRP-SF scaffolds significantly increased from day 1 to day 14 (***p < 0.001). The hOBs monocultured in the HRP-SF/dTCP and HRP-SF/TCP scaffolds presented superior ALP activity levels, as compared to those of hACs monocultured in the HRP-SF scaffolds. However, the ALP activity levels in these hybrid scaffolds were stable until day 7, and significantly higher from day 1 to day 14 (*p < 0.05 on HRP-SF/dTCP; **p < 0.01 on HRP-SF/TCP), and from day 7 to day 14 (*p < 0.05 on HRP-SF/TCP). The GAGs production was quantified in all types of culture (Figure 7c). The GAGs accumulation in the co-culture systems with hOBs and hACs, showed a significant increase in the BTCP constructs, from day 1 to day 14 and from day 7 to day 14 (*p < 0.05). Furthermore, no significant differences in GAGs production were found between the BdTCP and BTCP constructs at the same co-culture periods. In the HRP-SF scaffolds, there was a significant increase of GAGs accumulation after 14 days of culture with hACs (***p < 0.001). The HRP-SF/dTCP and HRP-SF/TCP scaffolds monocultured with hOBs did not show a significant increase of GAGs accumulation over the culture period. Nevertheless, the GAGs production on the HRP-SF/dTCP constructs at day 7, was significantly higher (*p < 0.05) as compared to that observed on the HRP-SF/TCP constructs.


21


Page 22 of 51

Figure 7. Quantification of in vitro osteogenesis and chondrogenesis. (a) Number of cells quantification based on DNA test, (b) normalized ALP activity against the number of cells, and (c) GAG‘s content on the BdTCP and BTCP co-cultured with hOBs and hACs for the period of 1, 7 and 14 days. Monocultures of hACs in the HRP-SF scaffolds and hOBs in the HRPSF/dTCP and HRP-SF/TCP scaffolds were used as controls. 3.4.3. Osteogenic and chondrogenic phenotype evaluation Histological and immunofluorescence analysis was performed to assess the phenotypic expression and activity of the osteoblasts and chondrocytes co-cultured in the BdTCP (Figure 8) and BTCP (Figure 9) constructs. The results obtained from the H&E staining images, showed that the hOBs and hACs were able to proliferate up to 14 days of culture. At day 7 and day 14, the hACs attached to the macro-pores walls of the HRP-SF layers and formed the typical self-aggregated clusters, whereas the hOBs spread and filled the inner porous of the HRP-SF/dTCP and HRP-SF/TCP layers. The newly formed ECM was stained with Sirius red, showing after 14 days of culture a well pronounced collagen matrix deposited in the co-cultured BdTCP and BTCP scaffolds. In the HRP-SF layers, the collagen matrix was mainly evidenced in the hACs aggregates. The GAGs deposition on the BdTCP and BTCP constructs was observed at day 14, by the positive staining for safranin-O. An increase of the ECM mineralization was observed up to 14 days of culture in the HRP-SF/dTCP and HRP-SF/TCP layers, as compared to the lower staining intensity observed on the HRP-SF layers. Since the hACs tend to form thick self-aggregated clusters, the staining intensity in these clusters was considerably higher. Up to 14 days of culture, no detectable differences


22

Page 23 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


were observed in the type of ECM produced by the hOBs and hACs co-cultured in the BdTCP and BTCP constructs, with that produced on the corresponding monolayered control scaffolds (Figure 10). The immunofluorescence images show that the hOBs cultured at the HRP-SF/dTCP and HRP-SF/TCP layers, and in the monolayered HRP-SF/dTCP and HRP-SF/TCP scaffolds, were able to express the bone-specific glycoproteins OPN and OCN over the 14 days of culture. The hACs in the co-culture and monoculture systems expressed the chondrogenic-related marker ACAN and the agglomerated chondrocytes were found expressing Col X, a specific marker of hypertrophic cartilage tissue.

Figure 8. Histological and immunofluorescence analysis of the hOBs and hACs co-cultured in the BdTCP scaffolds for 1, 7 and 14 days. Standard H&E staining was used to evaluate cell distribution and ECM formation, Sirius red (red) staining was used for collagen visualization at the ECM, Safranin-O (red) staining was used to detect GAGs formation, and Alizarin red (red) staining was used to detect calcium deposition and ECM mineralization (scale bar: 200 μm). Representative immunofluorescence images of the osteogenic-related markers OPN (green) and OCN (red), chondrogenic-related marker ACAN (red), and Col X (green) as a marker of hypertrophic chondrocytes. Nuclei are stained in blue (scale bar: 100 μm). The red arrow indicates a stained area of ECM mineralization. The yellow arrows indicate the stained HRP-SF and HRPSF/dTCP layers.


23


Page 24 of 51

Figure 9. Histological and immunofluorescence analysis of the hOBs and hACs co-cultured in the BTCP scaffolds for 1, 7 and 14 days. Standard H&E staining was used to evaluate cell distribution and ECM formation, Sirius red (red) staining was used for collagen visualization at the ECM, Safranin-O (red) staining was used to detect GAGs formation, and Alizarin red (red) staining was used to detect calcium deposition and ECM mineralization (scale bar: 200 μm). Representative immunofluorescence images of the osteogenic-related markers OPN (green) and OCN (red), chondrogenic-related marker ACAN (red), and Col X (green) as marker of hypertrophic chondrocytes. Nuclei are stained in blue (scale bar: 100 μm). The red arrow indicates a stained area of ECM mineralization. The yellow arrows indicate the stained HRP-SF and HRP-SF/TCP layers.


24

Page 25 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 10. Histological and immunofluorescence analysis of the hACs monocultured in the HRP-SF scaffolds and hOBs monocultured in the HRP-SF/dTCP and HRP-SF/TCP scaffolds for 1, 7 and 14 days. Standard H&E staining was used to evaluate cell distribution and ECM formation, Sirius red (red) staining was used for collagen visualization at the ECM, Safranin-O (red) staining was used to detect GAGs formation, and Alizarin red (red) staining was used to detect calcium deposition and ECM mineralization (scale bar: 200 μm). Representative immunofluorescence images of the osteogenicrelated markers OPN (green) and OCN (red), chondrogenic-related marker ACAN (red), and Col X (green) as marker of hypertrophic chondrocytes. Nuclei are stained in blue (scale bar: 100 μm). The red arrows indicate a stained areas of ECM mineralization. The yellow arrows indicate the stained HRP-SF, HRP-SF/dTCP and HRP-SF/TCP scaffolds.


25


Page 26 of 51

3.4.4. Osteogenic and chondrogenic genotype evaluation Analysis of relative gene expression after 14 days of co-culture, revealed that the hACs and hOBs expressed the typical chondrogenic- and osteogenic-related markers when cultured in the respective layers (HRP-SF and HRP-SF/dTCP or HRPSF/TCP) of the bilayered scaffolds (Figure 11). The hOBs cultured on the HRP-SF/dTCP layer, showed no significant differences on OCN (Figure 11a) and OPN (Figure 11b) gene expression, as compared to those of HRP-SF/TCP layer. However, a significantly higher gene expression was determined for OCN and OPN on the interface of the BdTCP (**p < 0.01). The hACs cultured at the HRP-SF layers of the BdTCP and BTCP constructs showed no significant differences for ACAN (Figure 11c) and Sox-9 (Figure 11d) gene expression, however, the interface region and subchondral bone-like layers also showed significantly higher gene expression values on the BdTCP constructs (**p < 0.01). Col X transcript levels (Figure 11e), showed no significant differences when compared the different layers of the BdTCP and BTCP constructs. Interestingly, the interface region presented the higher relative gene expression for this major collagen of chondrocytes hypertrophy, as compared to the cartilage-like (HRP-SF) and subchondral bone-like (HRP-SF/dTCP and HRP-SF/TCP) layers.

Figure 11. Relative expression of chondrogenic- and osteogenic-related transcripts, namely (a) OCN, (b) OPN, (c) ACAN, (d) Sox-9, and (e) Col X, by the hACs and hOBs co-cultured on the corresponding HRP-SF layer, interface, and HRPSF/dTCP or HRP-SF/TCP layers of the BdTCP and BTCP scaffolds, for 14 days. Fold-changes in relative gene expression were calculated using the 2-ΔΔCt method.


26

Page 27 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


4. Discussion The natural OC tissue consists of two distinct layers, the cartilage and subchondral bone, integrating a stable interface region. In general, the cartilage layer is flexible and supportive, formed by mature chondrocytes entrapped in an avascular environment composed of water, collagen type II (Col II) and proteoglycans. On the other side, the subchondral bone layer is formed of several cell types, e.g., osteoblasts, osteoclasts, osteocytes, and bone marrow stromal cells, involved by a vascularized environment in a bony lamella and trabeculae 37. This complex region also contains an inorganic component of HAp and collagen type I

38.

The interface region that interconnects these two very distinct tissues is characterized by

intermediate properties of both cartilage and subchondral bone tissues. Therefore, in OC TE the real challenge remains at creating hierarchical and well-interconnected structures that simultaneously meet the structural, mechanical, and biological requirements for chondral and subchondral bone regeneration. Different reports have proposed bilayered scaffolds for OC TE applications 9, 39. However, only few proposed a combination of high-performance SF scaffolds and nanocomposites of SF/CaP 28, 40. Yan et al. 24 developed porous bilayered scaffolds with an integrated SF layer into a SF/nanoCaP layer for OC defects regeneration, showing a spatially controllable porosity distribution through the scaffolds, and a CaP confinement to the subchondral bone-like layer. These scaffolds promoted in vitro cell attachment, proliferation and osteogenic differentiation of rabbit bone marrow mesenchymal stromal cells (rbMSCs), as well as, in vivo cartilage regeneration in the top of SF layer, and subchondral bone ingrowth in the bottom SF/nanoCaP layer. The SF scaffolds have shown to closely mimic the structure and mechanical complexity of tissues, such as cartilage and bone 21, 40. Through this method, the scaffolds could achieve a controlled pore size, high interconnectivity, and adjustable mechanical properties, depending on the SF concentration and induced porosity. Moreover, the formation of a silk-II structure (β-sheet conformation) results in high stiffer scaffolds

34.

HRP-crosslinked SF scaffolds developed using salt-leaching and freeze-drying technologies have

previously shown combined features of highly porous and mechanically stable SF scaffolds 21, with the elastic behavior of soft enzymatically crosslinked SF hydrogels 41, for cartilage regeneration purposes. Based on our previous findings, the herein developed bilayered scaffolds resulted from interconnected HRP-SF and HRPSF/TCP hybrid scaffolds incorporating Zn and Sr-dopants, for specific OC regeneration applications. As observed in the SEM micrographs (Figure 1a-f) and micro-CT analysis (Figure 2a-f), the BdTCP and BTCP scaffolds presented similar porous microstructure on both HRP-SF cartilage-like layer and HRP-SF/dTCP or HRP-SF/TCP subchondral bone-like layers, respectively. A macro- and micro-porous structure was achieved, with interconnected structures from the cartilage to the underlying subchondral bone-like layers. In a previous study, Li et al. 13 proposed a biphasic scaffold composed of SF and SF-coated Sr-hardystonite-gahnite (SHG-SF respectively for cartilage and subchondral bone-like phases, showing the formation of large pores on both cartilage (100-120 µm) and bone (400-500 µm) parts, as well as, the presence of small pores


27


Page 28 of 51

(20-40 µm) in the highly-interconnected pore walls. It is well-established that in cartilage and bone TE fields, the scaffolds should contain a wide range of pore size and interconnectivity 42. For example, in cartilage TE, larger and interconnected pores (300-500 µm) are required to allow cell infiltration, and the formation of self-aggregated clusters and ECM, whereas smaller pores (< 50 µm) are essential to endure adequate cell adhesion and proliferation 43. On the other side, larger pores have been reported as essential for nutrients exchange, matrix mineralization, and vascularization in bone tissue regeneration 44.

The interface region (~500 μm thick) that separates the cartilage and subchondral bone-like layers (Figure 1b and e) holds

a significant biomechanical function in the native OC tissue by adapting to microinjuries

38.

Moreover, it creates the

connection between the two OC layers, allowing that the highly-vascularized trabeculae of the subchondral bone tissue share nutrients with the adjacent articular cartilage, and at the same time anchor the collagen fibrils found in the deep zone of articular tissue 8. EDS (Figure 1) and micro-CT (Figure 2c, f and h) analyses demonstrated the confinement of the introduced TCP powders to the interfaces of both BdTCP and control BTCP scaffolds and respective HRP-SF/dTCP and HRP-SF/TCP layers. Additionally, the gradual increase of scaffolds porosity (Figure 2g), and decrease in TCP powders distribution (Figure 2h) from the subchondral bone-like layers to the cartilage-like layers, confirmed that the interfaces of the bilayered scaffolds has both bone and cartilage-like properties. Previous studies, have shown the importance of incorporating CaPs on scaffolds performance for subchondral bone mineralization and regeneration 24, 28, 40. Moreover, it was also reported that hydrogel-CaP composite scaffolds allowed calcified cartilage-like matrix formation and promoted osteointegration, as part of an OC interface TE strategy 45. The obtained results can also be related to the increase of the mean wall thickness (Table 1) on the HRP-SF/dTCP and HRP-SF/TCP layers. It has been reported 21, that the salt particles are partially dissolved during the saltinduced precipitation of SF solution, and the final pores formed after salt-leaching did not contain the same size as the original NaCl particles. Thus, the presence of β-TCP and ion-doped β-TCP powders in the HRP-SF solution, may have affected the salt dissolution and SF precipitation 34. The mechanical properties of the scaffolds are one of the most important issues when designing scaffolds either for bone or cartilage TE purposes 42. In case of OC TE, the mechanical properties of the reconstructive scaffolds can represent an even bigger challenge. In the native OC tissue, articular cartilage provides mechanical resistance to the joint, whereas the subchondral bone is responsible for withstanding most of the strength considering its larger area and lower elasticity

37.

Moreover, the mechanical properties of the articular cartilage are dependent of the zone. For example, the compressive modulus of the superficial, middle and deep zones is 0.079 MPa, 2.1 MPa and 7.75 MPa, respectively. In the case of subchondral bone, the compressive modulus is higher than that of cartilage and can reach 5.7 GPa 46. Some studies have been producing different silk-based scaffolds with high strength and elasticity, for respectively bone and cartilage regeneration applications 21, 40. It was shown that the wet compressive modulus of the combined bilayered structures was ~0.4 MPa, similar


28

Page 29 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


to that of the controls pure SF scaffolds and SF/nanoCaP scaffolds 24. In comparison to our results, the mechanical properties of the scaffolds are enhanced due to the incorporation of the ceramic powders (ZnSr-β-TCP and β-TCP). The compressive modulus (Figure 3b) and compressive strength (Figure 3c) of the BdTCP and BTCP scaffolds were significantly higher than that of the monolayered HRP-SF control scaffolds, showing that the bilayered scaffolds possess stratified compressive properties due to the different mechanical properties of the cartilage and subchondral bone-like layers. These properties can closely mimic the mechanical transitions existing at the different segments of the OC tissue 1. Interestingly, when looking at the mechanical behavior of the bilayered scaffolds, the BdTCP presented a significantly higher compressive strength than the BTCP. In fact, it have been reported that ion-dopants (e.g., Zn and Sr) can positively influence the mechanical properties of scaffolds through the modification of the lattice structure, microstructure, crystallinity and dissolution rate of the biomaterials 47. Despite the mechanical properties of the bilayered scaffolds being closer to that of the native human articular cartilage tissue (1.16 MPa in the superficial articular layer)

46

and not comparable to those of the human subchondral bone (1.15 GPa) 48,

different authors have demonstrated that temporary bone substitutes scaffolding should contain a compressive modulus of only few MPa

49-50.

This should be sufficient to support the initial regenerative process, e.g., withstand the hydrostatic

pressures and allow cellularity for ECM ingrowth and mineralization, and not equal to the mechanical strength of the native bone tissue. The storage modulus of the bilayered and monolayered scaffolds (Figure 3d) followed the same pattern of that determined for the static compressive modulus (Figure 3b). As demonstrated by Yan et al. 24, the BdTCP and BTCP scaffolds were able to maintain their integrity at high-frequency loading, due to the strong binding strength established between the interconnected HRP-SF layers and HRP-SF/dTCP or HRP-SF/TCP layers, respectively. The high loss factor (Figure 3e) showed the viscoelastic nature of both BdTCP (ranging from 0.58 to 0.52) and BTCP (ranging from 0.81 to 0.75) scaffolds, confirming that the proposed scaffolds can recover from high physical loads and easily adjust to fill the OC defect site (Movie S1). Despite the suitable mechanical properties of bilayered structures, understand if and how the osteochondral scaffolds can fail at the interface is an important issue to evaluate the mechanical integration between the layers. Holmes et al. 51, have seen that printed 3D biphasic scaffolds, presented not only improved mechanical compression results, as compared to the homogeneous control model, as well as, superior shear fracture results indicative of better interfacial integration. The degradation of the scaffolds is an important concern for all TE scenarios. Ideally, the scaffolds degradation should be adjusted to follow the regenerative process of the implanted in vivo tissues, which forces an equilibrium between their mechanical properties and degradation rate 52. When in contact to the body fluids, complex enzymatic pathways will guide the scaffolds degradation, and for that reason its in vitro enzymatic degradation is the best option to better predict the in vivo stability

53.

In this study, protease XIV degradation solutions were used based on previous studies proposing SF-based


29


Page 30 of 51

bilayered scaffolds for OC regeneration applications 24, 36. Even though this enzyme has no activity in the human body, the response of SF to the non-specific proteolytic degradation with protease XIV will mimic the in vivo simultaneous activity of several ECM modulatory enzymes 36. Our results showed that at different concentrations of protease XIV (Figure 4a), the BdTCP and BTCP scaffolds presented a slower degradation rate than that of the controls HRP-SF, HRP-SF/dTCP and HRPSF/TCP scaffolds. These differences can be attributed to the existence of a continuous interface region in the bilayered scaffolds, strongly integrating the two layers and hindering the proteolytic cleavage process induced by the protease XIV. Moreover, the incorporation of TCP phases into the SF polymer under enzymatic crosslinking, also made the structures less available for proteolytic degradation, especially at the highest protease XIV concentration (2 U/mL). It is well reported in the literature that the degradation of a polymer/CaP composite depends on their physicochemical properties, such as the polymer molecular weight and level of crosslinking, CaP phase, crystallinity, and surface area, as well as, the way they were integrated into the scaffolds 54. A previous study of our group 24, showed that pure silk scaffolds presented a superior weight loss profile than that incorporating nanoCaP, due to the dissolution of CaP. On the other hand, the combination of these two layers to form bilayered scaffolds seems to create a more complex structure with a degradation rate in between that of the controls scaffolds. The authors also observed that the bilayered silk/silk-nanoCaP scaffolds presented ~25% weight loss after 7 days in protease XIV degradation solution (0.0035 U/mL). Compared to our results, slower degradation rates were observed on the BdTCP (~10% weight loss) and BTCP (~20% weight loss) scaffolds, due to the strong covalent bonds between the tyrosine groups in SF enzymatic crosslinking process, that may have hindered the β-TCP dissolution and the degradation process 41, 55. At the molecular level, this may be an indication that the crosslinking level of SF and stoichiometry of β-TCP particles were adjusted in a way that mimic bone tissue at the nano-scale level 56. The superior mechanical properties observed on the bilayered structures and monolayered hybrid scaffolds (Figure 3b-e) also correlates to their resistance to enzymatic degradation. The in vitro bioactivity test performed to evaluate the biomineralization of the bilayered scaffolds, showed that the HRPSF/dTCP (Figure 4bii) and HRP-SF/TCP (Figure 4biv) layers induced the formation of apatite-like crystals after 15 days (Figure 4b) and 30 days (Figure S4) of soaking in SBF solution. No influence was observed from the presence of TCP (ZnSrβ-TCP and β-TCP) powders on the HRP-SF layers (Figure 4bi and iii), since no crystals were formed on these layers even after 30 days of immersion. These results suggest that the proposed bilayered scaffolds can be fastly integrated into an OC defect, using the subchondral bone-like layers as anchor. The importance of the bilayered 3D scaffold architectures for OC TE has been highlighted by numerous authors, stating that the interactions existing between the different cell types which compose the articular cartilage and subchondral bone tissue, can induce the diffusion of several paracrine signaling pathways that dictate the OC tissue regeneration 57. Thus, a hOBs/hACs


30

Page 31 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


based co-culture method was used in this study to investigate cell behavior in the complete OC grafts, BdTCP and BTCP scaffolds. In parallel, monocultures of hACs and hOBs on the respective monolayered control scaffolds (HRP-SF, HRPSF/dTCP and HRP-SF/TCP) were used as controls (Figure 1b). The bilayered scaffolds showed an outstanding performance for cell adhesion and viability, as observed from the metabolic activity (Figure 5) and SEM (Figure 6) results. On both coculture and monoculture systems, the metabolic activity of cells increased at significant levels over the culture period, which can be related to the chemical properties of SF, the crystalline structure of β-TCP and to the presence of ion-dopants, as well as, by the porous structure of the scaffolds changing the scaffolds pore size

58-59.

23.

As reported somewhere, different cell behaviors can be achieved only by

The scaffolds micro-porosity is responsible for cell attachment and spreading in the

first culture periods, whereas larger pores will allow cell infiltration, proliferation and subsequent ECM formation. These reports are in good agreement with our SEM results (Figure 6), showing from the first culture period the hOBs and hACs adhered and spread on the scaffolds surface, and fully infiltrated the interconnected large pores over the time-points. In addition, a significant increase of the cell number was detected up to 14 days in both monoculture and co-culture conditions (Figure 7a). The obtained results also showed that the hACs monocultured on the HRP-SF scaffolds presented the higher metabolic activity and proliferation rate, as compared to the HRP-SF/dTCP and HRP-SF/TCP monoculture systems. Interestingly, the ALP activity results (Figure 7b) showed an opposite trend with higher activity levels on the HRP-SF/dTCP and HRP-SF/TCP scaffolds monocultured with hOBs, suggesting that the ALP activity detected on the bilayered constructs can be mostly attributed to the hOBs in the co-culture systems. Considering that ALP is an important marker of osteogenic activity and is expressed by proliferating osteoblasts during in vitro osteogenesis 60, this is in agreement with the obtained results. These data can also be related to the use of the TCP (ZnSr-β-TCP and β-TCP) powders within the HRP-SF/dTCP and HRP-SF/TCP scaffolds, as observed from EDS (Figure 1c and f) and micro-CT (Figure 2c, f and h) analysis. Moreover, the apatite crystals formation after immersing in SBF solution (Figure 4bii, iv, vi and vii), also corroborates to this hypothesis. The positive influence of including CaPs into silk scaffolds, on the ALP activity expressed by BMSCs and human adiposederived stem cells (hASCs) is well known

24, 28.

The hACs monocultured on the HRP-SF scaffolds, showed significantly

higher GAGs content at day 14 (Figure 7c), which in this case can be explained by the capability of such cells to produce GAGs as one of the main components of the articular cartilage tissue 2. The histological observations confirmed that cells were able to adhere and deeply infiltrate into the highly interconnected porous scaffolds over the 14 days of monoculture and co-culture (Figures 8, 9, and 10). Interestingly, the monocultured and co-cultured hACs formed cell aggregates typically found in 3D cultures 61. An increase of the collagen deposition was also observed over the 14 days of culture in all tested groups, which is in accordance with the literature stating that both osteoblasts and chondrocytes cells are responsible for the collagen matrix deposition during bone and cartilage formation 62. Consistent


31


Page 32 of 51

with the results obtained from GAGs quantification (Figure 7c), a positive staining for safranin-O was observed after 14 days of culturing hACs on the HRP-SF scaffolds. Moreover, the HRP-SF/dTCP and HRP-SF/TCP layers also showed GAGs deposition at the newly-formed matrix. Previously, Jiang et al. 63 proposed a sequential co-culturing model using osteoblasts and chondrocytes in 3D culture showing that not only the individual chondrocytes were able to synthesize GAGs, as a GAGs matrix was formed in the co-culture system. Through the protein expression of the osteogenic-related markers, OPN and OCN, it was possible to confirm that the hOBs were present in the HRP-SF/dTCP and HRP-SF/TCP layers (Figures 8 and 9) and scaffolds (Figure 10). These are important osteogenic markers that regulate biomineralization 64, which is in agreement with the ECM mineralization observed after 14 days of culture in the same hybrid structures. Consistent with these observations, is also the superior relative gene expression obtained for OCN (Figure 11a) and OPN (Figure 11b) on the subchondral bone-like layers. The self-aggregated hACs cultured on the HRP-SF constructs and layers also showed high protein expression levels for the chondrogenic-related marker ACAN, accompanied by a superior relative gene expression for the same marker (Figure 11c) and Sox-9 (Figure 11d). However, the protein identification of Col X in the HRP-SF samples, also suggests pre-hypertrophic chondrocytes formation 65. This assumption was supported by the relative expression levels of Col X also identified on the interface region of the bilayered constructs (Figure 11e), especially on the BdTCP constructs which showed significantly higher gene expression levels of OCN and OPN at the interface, as markers of chondrocytes hypertrophy 66. In general, the BdTCP constructs showed superior influence over gene expression profile of osteogenic- and chondrogenic-related markers, as compared to the pure bilayered constructs, indicating that the incorporation of Zn and Sr in the β-TCP can improve hOBs osteogenic activity and affect hACs profile in the co-culture system. The osteogenic potential of Zn and Sr ions was previously reported in different studies 28, 67. For example, Pina et al. 28 observed that combining Zn and Sr within SF/β-TCP scaffolds increased hASCs osteogenic differentiation when compared to the single ion-doped constructs. The obtained results so far, have shown that the proposed biomaterials and combined technologies can be of great interest for producing biofunctional hierarchical scaffolds for complex tissues regeneration, particularly for OC TE applications. The established co-culture system showed the possibility of maintaining for long-term the co-culture of osteoblasts and chondrocytes in 3D model. This is especially relevant and a step forward if we consider the limitations of using chondrocytes in 2D culture. With this system, it was possible to test the outcome of the co-culture conditions and evaluate the expression of Col X as a specific interfacial marker. Nevertheless, the hACs differentiation into pre-hypertrophic chondrocytes and ultimately osteoblasts and osteocytes is a possibility, together with the whole structure mineralization in long-term co-cultures 65, 68.

To prevent this, a precise and controllable dual-chamber bioreactor system can be used to specifically deliver growth

factors required for tuning the behavior of pre-cultured osteoblasts and chondrocytes on the respective layers of the bilayered


32

Page 33 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


scaffolds. The evaluation of in vivo OC regeneration potential by the produced bilayered scaffolds is also an important point to be further meet.

5. Conclusions In this study, novel bilayered hybrid scaffolds made of a HRP-crosslinked SF (HRP-SF) cartilage-like layer fully integrating HRP-SF/ZnSr-β-TCP (HRP-SF/dTCP) and HRP-SF/β-TCP (HRP-SF/TCP) subchondral bone-like layers, are proposed for OC tissue regeneration. These scaffolds presented an adequate structural integrity, as well as, controllable porosity and TCP distribution alongside the scaffolds. Superior mechanical properties were observed on the bilayered scaffolds, with special relevance for the ion-doped bilayered structures. The in vitro co-culturing of human osteoblasts (hOBs) and human articular chondrocytes (hACs), showed cell adhesion, proliferation, and ECM production in the bilayered structures. The osteogenic activity on the HRP-SF/dTCP and HRP-SF/TCP layers was represented by the formation of mineralized matrix, whereas the chondrogenic inducement at the HRP-SF porous layers was characterized by GAGs deposition. The immunodetection and gene expression profile of chondrogenic- and osteogenic-related markers confirmed the presence of hACs and hOBs in the respective cartilage and subchondral bone-like layers of the bilayered scaffolds. Protein and genotypic expression levels of Col X indicated that that the proposed co-culture system may have the potential to induce chondrocytes pre-hypertrophy, with special influence from the ion-doping presence on the bilayered scaffolds. Although a complementary in vivo evaluation is necessary to fully validate these structures, and confirm the welfare of the ion presence, the physicochemical and biological properties observed herein suggest that these hierarchical scaffolds are encouraging candidates for OC tissue repair and regeneration.


33


Page 34 of 51

Author Contributions VPR performed the fabrication of the scaffolds, degradation profile analysis, in vitro bioactivity study with SBF, biological experiments, statistical analysis and prepared the manuscript. SP helped on scaffolds fabrication, performed XRD analysis, mechanical properties tests, in vitro bioactivity study with SBF and revised the manuscript. JBC performed the dynamic mechanical analysis. IFC performed the micro-CT analysis. LG-F helped on scaffolds fabrication and its characterization. MMF-G helped on the biological experiments. OCP contributed with reagents supply. ALO, JS-R, JMO and RLR defined the experimental work and revised the prepared manuscript.

Funding Sources This study was funded by the Portuguese Foundation for Science and Technology (FCT) for the HierarchiTech project (MERA-NET/0001/2014). The project FROnTHERA (NORTE-01-0145-FEDER-000023), supported by Norte Portugal Regional Operational Programme (NORTE 2020), under the PORTUGAL 2020 Partnership Agreement, through the European Regional Development Fund (ERDF). The FCT distinctions attributed to JMO (IF/00423/2012 and IF/01285/2015), and ALO (IF/00411/2013). VPR (PD/BD/113806/2015) and JBC (PD/BD/113803/2015) were awarded PhD scholarships under the financial support from FCT/MCTES and FSE/POCH, PD/59/2013. IFC was awarded a FCT PhD scholarship (SFRH/BD/99555/2014).

Supporting Information The Supporting Information is available free of charge on the ACS Publications website at DOI:xx.xxxx/. Immunofluorescence analysis of negative controls, relative gene expression data of individually cultured cells, low magnification SEM micrographs of bilayered scaffolds, and in vitro bioactivity evaluation of scaffolds after 30 days in SBF solution (PDF) Movie S1. Illustrative movie showing the resilient properties of the bilayered scaffolds (MPEG-1)

Author Information *E-mail: Viviana P. Ribeiro ([email protected]) E-mail: Sandra Pina ([email protected]) E-mail: João B. Costa ([email protected]) E-mail: Ibrahim Fatih Cengiz ([email protected]) E-mail: Luis García-Fernandez ([email protected]) E-mail: Maria del Mar Fernández-Gutiérrez ([email protected]) E-mail: Olga C. Paiva ([email protected]) E-mail: Ana L. Oliveira ([email protected]) E-mail: Julio San-Román ([email protected]) E-mail: Joaquim M. Oliveira ([email protected]) E-mail: Rui L. Reis ([email protected])


34

Page 35 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ORCID Sandra Pina (0000-0002-4361-1253) João B. Costa (0000-0002-2721-723X) Ibrahim Fatih Cengiz (0000-0003-0886-632X) Luis García-Fernandez (0000-0002-4179-2556) Maria del Mar Fernández-Gutiérrez (0000-0003-1739-1987) Ana L. Oliveira (0000-0001-8012-420) Joaquim M. Oliveira (0000-0001-7052-8837) Rui L. Reis (0000-0002-4295-6129)

Acknowledgments The authors thank to Raquel de Roba and Rosa Ana Ramirez (ICTP-CSIC) for the assistance in the cell culture laboratory, to Dr. Juan Parra Cáceres (Biomedical Research Networking Center in Bioengineering, Biomaterials and Nanomedicine (CIBER-BBN, ISCIII), and Teresa Oliveira for the assistance with the histological samples preparation and processing.

References (1) Yoo, J.; Yeo, E. D.; Lee, Y. K. Treatment of Osteochondral Lesions of the Talus in Athletes. The Korean Journal of Sports Medicine 2017, 35 (2), 77-85. (2) Dall'Oca, C.; Cengarle, M.; Costanzo, A.; Giannini, N.; Vacchiano, A.; Magnan, B. Current Concepts in Teatment of Early Knee Osteoarthritis and Osteochondral Lesions: The Role of Biological Augmentations. Acta Bio Medica Atenei Parmensis 2017, 88 (4-S), 5-10. (3) Meng, Z.; Huang, R. Topical Treatment of Degenerative Knee Osteoarthritis. Am J Med Sci. 2017, 355 (1), 6-12. (4) Wei, W.; Lambach, B.; Jia, G.; Flanigan, D.; Chaudhari, A. M.; Wei, L.; Rogers, A.; Payne, J.; Siston, R. A.; Knopp, M. V. Assessing the Effect of Football Play on Knee Articular Cartilage Using Delayed Gadolinium-Enhanced Mri of Cartilage (dGEMRIC). Magn. Reson. Imaging 2017, 39, 149-156. (5) Richter, D. L.; Schenck Jr, R. C.; Wascher, D. C.; Treme, G. Knee Articular Cartilage Repair and Restoration Techniques: A Review of the Literature. Sports health 2016, 8 (2), 153-160. (6) Frank, R. M.; Lee, S.; Levy, D.; Poland, S.; Smith, M.; Scalise, N.; Cvetanovich, G. L.; Cole, B. J. Osteochondral Allograft Transplantation of the Knee: Analysis of Failures at 5 Years. Am J Sports Med. 2017, 45 (4), 864-874. (7) Perdisa, F.; Sessa, A.; Filardo, G.; Marcacci, M.; Kon, E. Cell-Free Scaffolds for the Treatment of Chondral and Osteochondral Lesions. In BioOrthopaedics, Springer 2017, 139-149. (8) Yang, P. J.; Temenoff, J. S. Engineering Orthopedic Tissue Interfaces. Tissue Eng Part B Rev. 2009, 15 (2), 127-141. (9) Oliveira, J. M.; Rodrigues, M. T.; Silva, S. S.; Malafaya, P. B.; Gomes, M. E.; Viegas, C. A.; Dias, I. R.; Azevedo, J. T.; Mano, J. F.; Reis, R. L. Novel Kydroxyapatite/Chitosan Bilayered Scaffold for Osteochondral Tissue-Engineering Applications: Scaffold Design and Its Performance When Seeded with Goat Bone Marrow Stromal Cells. Biomaterials 2006, 27 (36), 6123-6137. (10) Yousefi, A. M.; Hoque, M. E.; Prasad, R. G.; Uth, N. Current Strategies in Multiphasic Scaffold Design for Osteochondral Tissue Engineering: A Review. J Biomed Mater Res A. 2015, 103 (7), 2460-2481. (11) Zhang, W.; Chen, J.; Tao, J.; Hu, C.; Chen, L.; Zhao, H.; Xu, G.; Heng, B. C.; Ouyang, H. W. The Promotion of Osteochondral Repair by Combined IntraArticular Injection of Parathyroid Hormone-Related Protein and Implantation of a Bi-Layer Collagen-Silk Scaffold. Biomaterials 2013, 34 (25), 6046-6057. (12) Guo, X.; Park, H.; Liu, G.; Liu, W.; Cao, Y.; Tabata, Y.; Kasper, F. K.; Mikos, A. G. In Vitro Generation of an Osteochondral Construct Using Injectable Hydrogel Composites Encapsulating Rabbit Marrow Mesenchymal Stem Cells. Biomaterials 2009, 30 (14), 2741-2752. (13) Li, J. J.; Kim, K.; Roohani-Esfahani, S.-I.; Guo, J.; Kaplan, D. L.; Zreiqat, H. A Biphasic Scaffold Based on Silk and Bioactive Ceramic with Stratified Properties for Osteochondral Tissue Regeneration. J Mater Chem B. 2015, 3 (26), 5361-5376. (14) Ding, X.; Zhu, M.; Xu, B.; Zhang, J.; Zhao, Y.; Ji, S.; Wang, L.; Wang, L.; Li, X.; Kong, D. Integrated Trilayered Silk Fibroin Scaffold for Osteochondral Differentiation of Adipose-Derived Stem Cells. ACS Appl Mater Interfaces. 2014, 6 (19), 16696-16705. (15) Mollet, B. B.; Comellas-Aragones, M.; Spiering, A.; Söntjens, S. H.; Meijer, E.; Dankers, P. Y. A Modular Approach to Easily Processable Supramolecular Bilayered Scaffolds with Tailorable Properties. J Mater Chem B. 2014, 2 (17), 2483-2493. (16) Kon, E.; Filardo, G.; Perdisa, F.; Venieri, G.; Marcacci, M. Acellular Matrix–Based Cartilage Regeneration Techniques for Osteochondral Repair. Oper. Tech. Orthop. 2014, 24 (1), 14-18. (17) Filardo, G.; Kon, E.; Di Martino, A.; Busacca, M.; Altadonna, G.; Marcacci, M. Treatment of Knee Osteochondritis Dissecans with a Cell-Free Biomimetic Osteochondral Scaffold: Clinical and Imaging Evaluation at 2-Year Follow-Up. Am J Sports Med. 2013, 41 (8), 1786-1793. (18) Bekkers, J. E.; Bartels, L. W.; Vincken, K. L.; Dhert, W. J.; Creemers, L. B.; Saris, D. B. Articular Cartilage Evaluation after Trufit Plug Implantation Analyzed by Delayed Gadolinium-Enhanced Mri of Cartilage (Dgemric). Am J Sports Med. 2013, 41 (6), 1290-1295. (19) Li, X.; Ding, J.; Wang, J.; Zhuang, X.; Chen, X. Biomimetic Biphasic Scaffolds for Osteochondral Defect Repair. Regen Biomater. 2015, 2 (3), 221-228. (20) Ribeiro, V. P.; Silva-Correia, J.; Nascimento, A. I.; da Silva Morais, A.; Marques, A. P.; Ribeiro, A. S.; Silva, C. J.; Bonifácio, G.; Sousa, R. A.; Oliveira, J. M. Silk-Based Anisotropical 3d Biotextiles for Bone Regeneration. Biomaterials 2017, 123, 92-106. (21) Yan, L.-P.; Oliveira, J. M.; Oliveira, A. L.; Caridade, S. G.; Mano, J. F.; Reis, R. L. Macro/Microporous Silk Fibroin Scaffolds with Potential for Articular Cartilage and Meniscus Tissue Engineering Applications. Acta Biomater. 2012, 8 (1), 289-301. (22) Zhao, Y. N.; Fan, J. J.; Li, Z. Q.; Liu, Y. W.; Wu, Y. P.; Liu, J. Effects of Pore Size on the Osteoconductivity and Mechanical Properties of Calcium Phosphate Cement in a Rabbit Model. Artif. Organs 2017, 41 (2), 199-204.


35


Page 36 of 51

(23) Farokhi, M.; Mottaghitalab, F.; Samani, S.; Shokrgozar, M. A.; Kundu, S. C.; Reis, R. L.; Fattahi, Y.; Kaplan, D. L. Silk Fibroin/Hydroxyapatite Composites for Bone Tissue Engineering. Biotechnol. Adv. 2017, 36 (1), 69-91. (24) Yan, L.-P.; Silva-Correia, J.; Oliveira, M. B.; Vilela, C.; Pereira, H.; Sousa, R. A.; Mano, J. F.; Oliveira, A. L.; Oliveira, J. M.; Reis, R. L. Bilayered Silk/SilkNanocap Scaffolds for Osteochondral Tissue Engineering: In Vitro and in Vivo Assessment of Biological Performance. Acta Biomater. 2015, 12, 227-241. (25) Ogose, A.; Hotta, T.; Hatano, H.; Kawashima, H.; Tokunaga, K.; Endo, N.; Umezu, H. Histological Examination of Β-Tricalcium Phosphate Graft in Human Femur. J Biomed Mater Res A. 2002, 63 (5), 601-604. (26) Takahashi, Y.; Yamamoto, M.; Tabata, Y. Osteogenic Differentiation of Mesenchymal Stem Cells in Biodegradable Sponges Composed of Gelatin and Β-Tricalcium Phosphate. Biomaterials 2005, 26 (17), 3587-3596. (27) Yanoso-Scholl, L.; Jacobson, J. A.; Bradica, G.; Lerner, A. L.; O'Keefe, R. J.; Schwarz, E. M.; Zuscik, M. J.; Awad, H. A. Evaluation of Dense Polylactic Acid/Beta-Tricalcium Phosphate Scaffolds for Bone Tissue Engineering. J Biomed Mater Res A. 2010, 95 (3), 717-726. (28) Pina, S.; Canadas, R.; Jiménez, G.; Perán, M.; Marchal, J.; Reis, R.; Oliveira, J. Biofunctional Ionic-Doped Calcium Phosphates: Silk Fibroin Composites for Bone Tissue Engineering Scaffolding. Cells Tissues Organs 2017, 204, 150-163. (29) Hunt, J. A.; Chen, R.; van Veen, T.; Bryan, N. Hydrogels for Tissue Engineering and Regenerative Medicine. J Mater Chem B. 2014, 2 (33), 5319-5338. (30) Liao, J.; Tian, T.; Shi, S.; Xie, X.; Ma, Q.; Li, G.; Lin, Y. The Fabrication of Biomimetic Biphasic Can-Pac Hydrogel with a Seamless Interfacial Layer Applied in Osteochondral Defect Repair. Bone research 2017, 5, 17018. (31) Ribeiro, V. P.; da Silva Morais, A.; Maia, F. R.; Canadas, R. F.; Costa, J. B.; Oliveira, A. L.; Oliveira, J. M.; Reis, R. L. Combinatory Approach for Developing Silk Fibroin Scaffolds for Cartilage Regeneration. Acta Biomater. 2018, 72, 197-181. (32) Horan, R. L.; Antle, K.; Collette, A. L.; Wang, Y.; Huang, J.; Moreau, J. E.; Volloch, V.; Kaplan, D. L.; Altman, G. H. In Vitro Degradation of Silk Fibroin. Biomaterials 2005, 26 (17), 3385-3393. (33) Tas, A. C. Synthesis of Biomimetic Ca-Hydroxyapatite Powders at 37 C in Synthetic Body Fluids. Biomaterials 2000, 21 (14), 1429-1438. (34) Kim, U.-J.; Park, J.; Kim, H. J.; Wada, M.; Kaplan, D. L. Three-Dimensional Aqueous-Derived Biomaterial Scaffolds from Silk Fibroin. Biomaterials 2005, 26 (15), 2775-2785. (35) Kannan, S.; Goetz-Neunhoeffer, F.; Neubauer, J.; Ferreira, J. M. Synthesis and Structure Refinement of Zinc-Doped Β-Tricalcium Phosphate Powders. J. Am. Ceram. Soc. 2009, 92 (7), 1592-1595. (36) Reardon, P. J.; Konwarh, R.; Knowles, J. C.; Mandal, B. B. Mimicking Hierarchical Complexity of the Osteochondral Interface Using Electrospun Silk– Bioactive Glass Composites. ACS Appl Mater Interfaces. 2017, 9 (9), 8000-8013. (37) Yan, L.-P.; Oliveira, J. M.; Oliveira, A. L.; Reis, R. L. Current Concepts and Challenges in Osteochondral Tissue Engineering and Regenerative Medicine. ACS Biomater Sci Eng. 2015, 1 (4), 183-200. (38) Madry, H.; van Dijk, C. N.; Mueller-Gerbl, M. The Basic Science of the Subchondral Bone. Knee Surg. Sports Traumatol. Arthrosc. 2010, 18 (4), 419433. (39) Pereira, D. R.; Canadas, R. F.; Silva-Correia, J.; Marques, A. P.; Reis, R. L.; Oliveira, J. M. Gellan Gum-Based Hydrogel Bilayered Scaffolds for Osteochondral Tissue Engineering. Key Eng. Mater. 2014, 587, 255-260. (40) Yan, L.-P.; Silva-Correia, J.; Correia, C.; Caridade, S. G.; Fernandes, E. M.; Sousa, R. A.; Mano, J. F.; Oliveira, J. M.; Oliveira, A. L.; Reis, R. L. Bioactive Macro/Micro Porous Silk Fibroin/Nano-Sized Calcium Phosphate Scaffolds with Potential for Bone-Tissue-Engineering Applications. Nanomedicine 2013, 8 (3), 359-378. (41) Yan, L.-P.; Silva-Correia, J.; Ribeiro, V. P.; Miranda-Gonçalves, V.; Correia, C.; da Silva Morais, A.; Sousa, R. A.; Reis, R. M.; Oliveira, A. L.; Oliveira, J. M. Tumor Growth Suppression Induced by Biomimetic Silk Fibroin Hydrogels. Sci. Rep. 2016, 6, 31037. (42) Puppi, D.; Chiellini, F.; Piras, A.; Chiellini, E. Polymeric Materials for Bone and Cartilage Repair. Prog. Polym. Sci. 2010, 35 (4), 403-440. (43) Griffon, D. J.; Sedighi, M. R.; Schaeffer, D. V.; Eurell, J. A.; Johnson, A. L. Chitosan Scaffolds: Interconnective Pore Size and Cartilage Engineering. Acta Biomater. 2006, 2 (3), 313-320. (44) Nazarov, R.; Jin, H.-J.; Kaplan, D. L. Porous 3-D Scaffolds from Regenerated Silk Fibroin. Biomacromolecules 2004, 5 (3), 718-726. (45) Khanarian, N. T.; Jiang, J.; Wan, L. Q.; Mow, V. C.; Lu, H. H. A Hydrogel-Mineral Composite Scaffold for Osteochondral Interface Tissue Engineering. Tissue Eng Part A. 2011, 18 (5-6), 533-545. (46) Chen, S.; Falcovitz, Y.; Schneiderman, R.; Maroudas, A.; Sah, R. Depth-Dependent Compressive Properties of Normal Aged Human Femoral Head Articular Cartilage: Relationship to Fixed Charge Density. Osteoarthritis Cartilage 2001, 9 (6), 561-569. (47) Bose, S.; Fielding, G.; Tarafder, S.; Bandyopadhyay, A. Trace Element Doping in Calcium Phosphate Ceramics to Understand Osteogenesis and Angiogenesis. Trends Biotechnol. 2013, 31 (10), 594-605. (48) Choi, K.; Kuhn, J. L.; Ciarelli, M. J.; Goldstein, S. A. The Elastic Moduli of Human Subchondral, Trabecular, and Cortical Bone Tissue and the SizeDependency of Cortical Bone Modulus. J. Biomech. 1990, 23 (11), 1103-1113. (49) Blaker, J. J.; Maquet, V.; Jérôme, R.; Boccaccini, A. R.; Nazhat, S. Mechanical Properties of Highly Porous Pdlla/Bioglass® Composite Foams as Scaffolds for Bone Tissue Engineering. Acta Biomater. 2005, 1 (6), 643-652. (50) Zhang, R.; Ma, P. X. Poly (Α-Hydroxyl Acids)/Hydroxyapatite Porous Composites for Bone-Tissue Engineering. I. Preparation and Morphology. J Biomed Mater Res A. 1999, 44 (4), 446-455. (51) Holmes, B.; Zhu, W.; Li, J.; Lee, J. D.; Zhang, L. G. Development of Novel Three-Dimensional Printed Scaffolds for Osteochondral Regeneration. Tissue Eng Part A. 2014, 21 (1-2), 403-415. (52) Makris, E. A.; Gomoll, A. H.; Malizos, K. N.; Hu, J. C.; Athanasiou, K. A. Repair and Tissue Engineering Techniques for Articular Cartilage. Nat Rev Rheumatol. 2015, 11 (1), 21-34. (53) Almeida, L. R.; Martins, A. R.; Fernandes, E. M.; Oliveira, M. B.; Correlo, V. M.; Pashkuleva, I.; Marques, A. P.; Ribeiro, A. S.; Durães, N. F.; Silva, C. J. New Biotextiles for Tissue Engineering: Development, Characterization and in Vitro Cellular Viability. Acta Biomater. 2013, 9 (9), 8167-8181. (54) Abadi, M.; Ghasemi, I.; Khavandi, A.; Shokrgozar, M.; Farokhi, M.; Homaeigohar, S. S.; Eslamifar, A. Synthesis of Nano Β-Tcp and the Effects on the Mechanical and Biological Properties of Β-Tcp/Hdpe/Uhmwpe Nanocomposites. Polymer Composites 2010, 31 (10), 1745-1753. (55) Yan, L. P.; Oliveira, J. M.; Oliveira, A. L.; Reis, R. L. Core-Shell Silk Hydrogels with Spatially Tuned Conformations as Drug-Delivery System. J. Tissue Eng. Regen. Med. 2016, 11 (11), 3168-3177. (56) Tamaddon, M.; Czernuszka, J. The Need for Hierarchical Scaffolds in Bone Tissue Engineering. Hard Tissue 2013, 2, 1-8. (57) O'Shea, T. M.; Miao, X. Bilayered Scaffolds for Osteochondral Tissue Engineering. Tissue Eng Part B Rev. 2008, 14 (4), 447-464. (58) Murphy, C. M.; Haugh, M. G.; O'Brien, F. J. The Effect of Mean Pore Size on Cell Attachment, Proliferation and Migration in Collagen– Glycosaminoglycan Scaffolds for Bone Tissue Engineering. Biomaterials 2010, 31 (3), 461-466. (59) Lien, S.-M.; Ko, L.-Y.; Huang, T.-J. Effect of Pore Size on Ecm Secretion and Cell Growth in Gelatin Scaffold for Articular Cartilage Tissue Engineering. Acta Biomater. 2009, 5 (2), 670-679. (60) Parekh, N.; Hushye, C.; Warunkar, S.; Gupta, S. S.; Nisal, A. In Vitro Study of Novel Microparticle Based Silk Fibroin Scaffold with Osteoblast-Like Cells for Load-Bearing Osteo-Regenerative Applications. RSC Advances 2017, 7 (43), 26551-26558.


36

Page 37 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(61) Correia, C. R.; Moreira-Teixeira, L. S.; Moroni, L.; Reis, R. L.; van Blitterswijk, C. A.; Karperien, M.; Mano, J. F. Chitosan Scaffolds Containing Hyaluronic Acid for Cartilage Tissue Engineering. Tissue Eng Part C Methods. 2011, 17 (7), 717-730. (62) Gómez-Picos, P.; Eames, B. F. On the Evolutionary Relationship between Chondrocytes and Osteoblasts. Frontiers in genetics 2015, 6, 297. (63) Jiang, J.; Nicoll, S. B.; Lu, H. H. Co-Culture of Osteoblasts and Chondrocytes Modulates Cellular Differentiation in Vitro. Biochem. Biophys. Res. Commun. 2005, 338 (2), 762-770. (64) Martins, A.; Alves da Silva, M. L.; Faria, S.; Marques, A. P.; Reis, R. L.; Neves, N. M. The Influence of Patterned Nanofiber Meshes on Human Mesenchymal Stem Cell Osteogenesis. Macromol. Biosci. 2011, 11 (7), 978-987. (65) Yang, L.; Tsang, K. Y.; Tang, H. C.; Chan, D.; Cheah, K. S. Hypertrophic Chondrocytes Can Become Osteoblasts and Osteocytes in Endochondral Bone Formation. Proc Natl Acad Sci USA. 2014, 111 (33), 12097-12102. (66) Gerstenfeld, L.; Shapiro, F. Expression of Bone-Specific Genes by Hypertrophic Chondrocytes: Implications of the Complex Functions of the Hypertrophic Chondrocyte During Endochondral Bone Development. J. Cell. Biochem. 1996, 62 (1), 1-9. (67) Place, E. S.; Rojo, L.; Gentleman, E.; Sardinha, J. P.; Stevens, M. M. Strontium-and Zinc-Alginate Hydrogels for Bone Tissue Engineering. Tissue Eng Part A. 2011, 17 (21-22), 2713-2722. (68) Zhou, X.; von der Mark, K.; Henry, S.; Norton, W.; Adams, H.; de Crombrugghe, B. Chondrocytes Transdifferentiate into Osteoblasts in Endochondral Bone During Development, Postnatal Growth and Fracture Healing in Mice. PLoS Genet. 2014, 10 (12), e1004820.


37


Page 38 of 51

Table of Contents Graphic

(3,3 cm x 1,25 cm) (TOC)


38

Page 39 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Scheme 1. Schematics of the experimental setup adopted in this study. (a) Methodology used for preparing the BdTCP and BTCP scaffolds, creating HRP-SF scaffolds as cartilage-like layers in combination with HRPSF/dTCP and HRP-SF/TCP scaffolds in an 80/20 (w/w) ratio as subchondral bone-like layers. (b) Temporal approach followed for co-culturing human osteoblasts (hOBs) and human articular chondrocytes (hACs) on the corresponding HRP-SF/dTCP, HRP-SF/TCP and HRP-SF layers of the BdTCP and BTCP scaffolds. Monoculture system of hOBs on the monolayered HRP-SF/dTCP and HRP-SF/TCP scaffolds, and hACs on the monolayered HRP-SF scaffold used as controls. Scaffolds were prepared with 8 mm diameter and height (HRP-SF layer: 3 mm in height; HRP-SF/dTCP and HRP-SF/TCP layers: 5 mm in height). 50x46mm (600 x 600 DPI)



Figure 1. SEM micrographs (scale bar: 100 μm) and EDS elemental analysis with the detected ionic elements and Ca/P ratios determined at the different regions of the BdTCP and BTCP scaffolds. (a, d) HRP-SF layers, (b, e) interface, (c) HRP-SF/dTCP layer, and (f) HRP-SF/TCP layer. 25x12mm (600 x 600 DPI)


Page 40 of 51

Page 41 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. Micro-CT analysis of the BdTCP and BTCP scaffolds. 2D images of the (a, d) HRP-SF layers, (b) HRP-SF/dTCP layer and (e) HRP-SF/TCP layer (scale bar: 2 mm). 3D reconstructions of the (c) BdTCP scaffolds and (f) BTCP scaffolds showing the HRP-SF matrix in brown and the HRP-SF/dTCP or HRP-SF/TCP matrices in blue (scale bar: 2 mm). Change of (g) 2D porosity distribution across the vertical axis of the scaffolds and (h) TCP (ZnSr-β-TCP and β-TCP) component distribution from the subchondral bone-like layers across the vertical axis of the scaffolds (i.e. HRP-SF/dTCP and HRP-SF/TCP, respectively) to the cartilage-like layers (HRP-SF). 25x20mm (600 x 600 DPI)



Figure 3. Crystalline structure and mechanical characterization of the BdTCP and BTCP scaffolds: (a) XRD patterns of the HRP-SF, HRP-SF/dTCP and HRP-SF/TCP scaffolds, and zoomed area of graph in a 2θ range from 27 to 36, showing the shift toward higher 2θ in the HRP-SF/dTCP scaffolds. (b) Compressive modulus and (c) compressive strength of the bilayered scaffolds (BdTCP and BTCP) and controls (HRP-SF, HRPSF/dTCP and HRP-SF/TCP), measured in the hydrated state. (d) Storage modulus (E’) and (e) loss factor (tan ��) of the bilayered scaffolds (BdTCP and BTCP) and controls (HRP-SF, HRP-SF/dTCP and HRPSF/TCP) obtained by DMA, tested at physiological conditions (pH 7.4 and 37ºC). 50x43mm (600 x 600 DPI)


Page 42 of 51

Page 43 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. Degradation profile and in vitro bioactivity evaluation of the BdTCP and BTCP scaffolds. (a) Enzymatic degradation of the bilayered scaffolds (BdTCP and BTCP) and controls (HRP-SF, HRP-SF/dTCP and HRP-SF/TCP), measured for a period of 30 days (% weight). Samples immersed in PBS were used as controls. (b) SEM micrographs (scale bar: 10 μm) and EDS elemental analysis with the detected ionic elements and Ca/P ratios determined at the (i, iii) HRP-SF layers, (ii) HRP-SF/dTCP layer and (iv) HRPSF/TCP layer of the BdTCP and BTCP scaffolds, after 15 days of immersion in SBF solution. (v) HRP-SF scaffolds, (vi) HRP-SF/dTCP scaffolds and (vii) HRP-SF/TCP scaffolds were also evaluated. 48x22mm (600 x 600 DPI)



Figure 5. Metabolic activity of the hOBs and hACs co-cultured in the BdTCP and BTCP scaffolds for 1, 7 and 14 days, determined by Alamar blue assay. Monocultures of hACs in the HRP-SF scaffolds and hOBs in the HRP-SF/dTCP and HRP-SF/TCP scaffolds were used as controls. 48x18mm (600 x 600 DPI)


Page 44 of 51

Page 45 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 6. SEM micrographs of the co-cultured hACs and hOBs for 1, 7 and 14 days, on the corresponding HRP-SF layer, HRP-SF/dTCP or HRP-SF/TCP layers of the BdTCP and BTCP scaffolds, respectively (scale bar: 100 μm). Monocultures of hACs in the monolayered HRP-SF scaffolds, and hOBs in the monolayered HRPSF/dTCP and HRP-SF/TCP scaffolds were observed as controls. The yellow arrows indicate the cells inside the macro-pores. 63x61mm (600 x 600 DPI)



Figure 7. Quantification of in vitro osteogenesis and chondrogenesis. (a) Number of cells quantification based on DNA test, (b) normalized ALP activity against the number of cells, and (c) GAG‘s content on the BdTCP and BTCP co-cultured with hOBs and hACs for the period of 1, 7 and 14 days. Monocultures of hACs in the HRP-SF scaffolds and hOBs in the HRP-SF/dTCP and HRP-SF/TCP scaffolds were used as controls. 32x36mm (600 x 600 DPI)


Page 46 of 51

Page 47 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 8. Histological and immunofluorescence analysis of the hOBs and hACs co-cultured in the BdTCP scaffolds for 1, 7 and 14 days. Standard H&E staining was used to evaluate cell distribution and ECM formation, Sirius red (red) staining was used for collagen visualization at the ECM, Safranin-O (red) staining was used to detect GAGs formation, and Alizarin red (red) staining was used to detect calcium deposition and ECM mineralization (scale bar: 200 μm). Representative immunofluorescence images of the osteogenicrelated markers OPN (green) and OCN (red), chondrogenic-related marker ACAN (red), and Col X (green) as a marker of hypertrophic chondrocytes. Nuclei are stained in blue (scale bar: 100 μm). The red arrow indicates a stained area of ECM mineralization. The yellow arrows indicate the stained HRP-SF and HRPSF/dTCP layers. 48x34mm (600 x 600 DPI)



Figure 9. Histological and immunofluorescence analysis of the hOBs and hACs co-cultured in the BTCP scaffolds for 1, 7 and 14 days. Standard H&E staining was used to evaluate cell distribution and ECM formation, Sirius red (red) staining was used for collagen visualization at the ECM, Safranin-O (red) staining was used to detect GAGs formation, and Alizarin red (red) staining was used to detect calcium deposition and ECM mineralization (scale bar: 200 μm). Representative immunofluorescence images of the osteogenicrelated markers OPN (green) and OCN (red), chondrogenic-related marker ACAN (red), and Col X (green) as marker of hypertrophic chondrocytes. Nuclei are stained in blue (scale bar: 100 μm). The red arrow indicates a stained area of ECM mineralization. The yellow arrows indicate the stained HRP-SF and HRPSF/TCP layers. 48x35mm (600 x 600 DPI)


Page 48 of 51

Page 49 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 10. Histological and immunofluorescence analysis of the hACs monocultured in the HRP-SF scaffolds and hOBs monocultured in the HRP-SF/dTCP and HRP-SF/TCP scaffolds for 1, 7 and 14 days. Standard H&E staining was used to evaluate cell distribution and ECM formation, Sirius red (red) staining was used for collagen visualization at the ECM, Safranin-O (red) staining was used to detect GAGs formation, and Alizarin red (red) staining was used to detect calcium deposition and ECM mineralization (scale bar: 200 μm). Representative immunofluorescence images of the osteogenic-related markers OPN (green) and OCN (red), chondrogenic-related marker ACAN (red), and Col X (green) as marker of hypertrophic chondrocytes. Nuclei are stained in blue (scale bar: 100 μm). The red arrows indicate a stained areas of ECM mineralization. The yellow arrows indicate the stained HRP-SF, HRP-SF/dTCP and HRP-SF/TCP scaffolds. 48x49mm (600 x 600 DPI)



Figure 11. Relative expression of chondrogenic- and osteogenic-related transcripts, namely (a) OCN, (b) OPN, (c) ACAN, (d) Sox-9, and (e) Col X, by the hACs and hOBs co-cultured on the corresponding HRP-SF layer, interface, and HRP-SF/dTCP or HRP-SF/TCP layers of the BdTCP and BTCP scaffolds, for 14 days. Foldchanges in relative gene expression were calculated using the 2^-ΔΔCt method. 48x35mm (600 x 600 DPI)


Page 50 of 51

Page 51 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(3,3 cm x 1,25 cm) (TOC) 32x12mm (600 x 600 DPI)


Enzymatically crosslinked silk fibroin-based hierarchical scaffolds for

Recommend Documents